FINN THIESEN A MANUAL OF CLASSICAL PERSIAN PROSODY with chapters on Urdu, Karakhanidic and Ottoman prosody 1982 OTTO HARRASSOWI I'Z • WIESBADEN CIP-Kurzlitelaufnahme dcr Dcutschen Bibliothek Thiesen, Finn : A manual of classical Persian prosody: with chapters on Urdu. Karakhanidic and Ottoman prosody / Finn Thiesen. - Wiesbaden: Harrassowitz. 1982. ISBN 3-447-02104-7 Publication of this book was aided by a grant from the Danish Research Council for the Humanities and by an auxiliary grant from the Iranian Culture Foundation. Calligraphy by Ali Alparslan, K. L. Gupta. Amanollah MQqcn, and the author. Alle Rechte vorbehalten. © Otto Harrassowitz. Wiesbaden 1982. Photomechanische und photographische Wiedergabe nur mit ausdriicklicher Genehmigung des Verlages. Gesamtherstellung: Imprimeric Orientalistc, Leuven. — Printed in Belgium. to jamAl rezaT without whose timely encouragement this work might not have been completed CORRIGENDA to Finn Thiesen A Manual of Classical Persian Prosody Page \iv\ line II: Page xxii, line 3: Page xxv, line 22: Page 64: Page 183-194 passim : Page 185. line 1 1 : For discribcd read For read For read Delete footnote 12, For hamza read For call read described hamza calls TABLE OP CONTENTS PREFACE XI LIST OF ABBREVIATIONS XIX BIBLIOGRAPHY XXI PART ONE: THE SYLLABLE 1 I The Vowel Systems of Classical and Modern Persian 3 II Transcription 8 HI Long and Short Vowels 11 IV Open and Closed Syllables 13 V Overlong Syllables 15 * VI Hiatus or hamze 19 m VII kasre ye ezafe and vdv e at/ 23 VIII Final Short Vowels 28 IX Nasalization .... 39 X Final Long Vowels 43 XI Altering the Syllabic Structure of Words by Suppressing Short Vowels, Doubling Consonants, etc 51 XU Poetic Licence 61 XIII The Arabic Element . . 69 XIV Rhyme 73 [Rhyme Arrangements (= Verse Forms)] 77 [emale] 81 Exercise 82 Appendix to Part One: List of Passages Quoted 84 PART TWO: THE RHYTHM 87 XV How to Ascertain the Rhythm of a Given Poem .... 89 XVI The Circles of the Prosodists 102 Exercise Ill XVII Metres with Three Syllables to the rokn (The First Circle) 1 12 bahr e motaqareb 112 hahr e motadarek 121 VIII TABLE OF CONTENTS XVI II Slow Metres (The Second Circle) 122 bahr e hazaj 122 bahr e ramai 126 bahr e rajaz 130 XIX Fast Metres (The Third Circle) 132 bahr e ramai e max him . . , 132 bahr e rajaz e matvi 137 bahr e surf V 138 [bahr e hazaj ] 140 XX Compound Metres (The Fourth Circle) 144 bahr e mojtass 144 bahr e xajif . . . ! 148 bahr e monsareh 151 bahr e mozdreS 153 bahr e qarib 158 bahr e moqtazab 160 XXI Compound Metres Traditionally Considered to be Varieties of Simple Metres (The Fifth Circle) 161 bahr e ramai e masku! 161 [bahr e rajaz] 162 [ bahr e hazaj] 163 XXII The robd 9/ Metre 166 XX II I Arabic Metres and Sundry Other Matters 174 [The Arrangement of a divan] 176 PART THREE: ADAPTATION INTO OTHER LANGUAGES 179 XXIV The Adaptation of Classical Persian Prosody to Urdu . . 181 [The Urdu Vowel and Consonant Systems) 182 [Transcription] 184 [Long and Short Vowels] 185 [Open and Closed Syllables] 186 [Overlong Syllables] 189 [Hiatus or hamza ] 191 [rav e Satf and izdja] 1 94 [Final Short Vowels] 194 [Nasalization] . 195 [Final Long Vowels] 197 [Altering the Syllabic Structure of Words] 203 [Poetic Licence] 205 [Rhythms] 209 TABLE OF CONTENTS IX XXV The Adaptation of Classical Persian Prosody to Kara- khanidie (Middle East-Turkish) 210 XXVI The Adaptation of Classical Persian Prosody to Ottoman Turkish 217 Exercise 225 appendix one: Reference List of Metres 227 Key to the bohur in Appendix One 256 appendix two: Reference List of Spellings 257 SUPPLEMENTARY NOTES 260 INDEX . . 265 APPENDIX three: The Arrangement of Moulavi’s divan . . . foil. p.274 PREFACE Students of Classical Persian often take up the subject in the hope that one day they will be able to relax in an armchair and enjoy the great classics at their leisure. In this they never succeed, though not because Classical Persian is a particularly difficult language. On the contrary, in spite of its consonantal script, it must be reckoned as a comparatively easy language. Neither is it because the student has to master a new culture together with the new language. This is a real difficulty, but not insurmountable 1 . The reason is that the Classical Persian books were never meant for armchair reading. The very word book had quite different connotations in those days. It was not a cheap mass-produced commodity, but a rarity and a luxury. Few could read one, still fewer could afford to possess one. In order to acquire a book one had to copy it oneself or to pay someone a month s wage to copy it. An alternative was to learn its contents wholly or partly by heart. Certainly the reader who paid so dearly for his book would not have been satisfied with a few hours’ light entertainment, and might prefer a work which could not be understood without effort. Ideally a literary Classical Persian work should be so beautiful in form and so rich in content that the reader would return to it again and again. Now masterpieces, outstanding both in form and content, are rare, for few indeed are those who have the vision and ideas needed to create a rich content, and these they must combine with formal mastery. However, formal mastery is to some extent a skill which may be acquired. So Classical Persian literature in time developed a tradition for formal perfection and has a surprising number of formal masters. Naturally, poetry was preferred to prose. The student must approach a Classical Persian literary work with a corresponding attitude. It has to be read again and again, and can be understood only after a prolonged but rewarding study. In order to understand 1 For each word the student has lo acquire a new and strange set of connotations, painfully and often through a process of trial and error. One must know that rain is pleasant and sunshine unpleasant. The owl is a symbol of imbecility. The verb to eat should automatically conjure up a picture of someone sitting on the floor eating with his fingers. Music is a vice and begging a virtue, and the very idea of God having a son is an abomination beyond words. xn PREFACE and appreciate Classical Persian poetry the student must strive to acquire the knowledge and skills of those learned and fastidious readers for whom the poems were intended. In order to enjoy the formal perfection — certainly no mean enjoyment — he must study Classical Persian rhetoric, poetics and prosody. Moreover a knowledge of these disciplines, especially prosody, can he very helpful in establishing correct readings and arriving at satisfactory interpretations. There is an obvious need for a general introduction to Classical Persian prosody, and when I began work upon this book eleven years ago my aim was no more than that. My first draft corresponds roughly to the main text of Chapters I-XV. However as work progressed, the scope of the book widened, and in its present form it will, I hope, serve not only as an introduction for students, but also as a reference work for Persian, Urdu and Turkish scholars, containing not only general information on prosody, but also a number of original discoveries in the fields of Classical Persian, Urdu and Turkish prosody. par i one expounds and discusses all the well-known rules of scansion that can be deduced from Classical Persian poetry. In its logical construction it is similar to a text-book of mathematics, in that each rule presupposes a knowledge of the preceding rules, but not of any succeeding rules. As far as possible the examples have been chosen so that they can be scanned with recourse exclusively to rules already given. In each case it has been shown that the rule is not a poetic idiosyncrasy, but has a phonetic explanation. Such explanations (which have not to my knowledge been attempted before) are mostly printed in a smaller fount. Taking the relevant scansion rules as the point of departure I have also shown the phonetic realities underlying the Persian concept of fluency and melodiousness of language 2 . In this research I had the benefit of the statistics compiled by L.P. Elwell-Sutton in his Persian Metres. The last chapter in Pari One deals with rhyme, which has received a functional treatment independent of the traditional approach. In order that so many unfamiliar rhythms should not divert the attention of the student from the basic mechanisms of scansion, I decided that all the examples for the rules in Part One should be in one and the same rhythm. For further convenience, I wanted the examples to be in a uniform style, as far as possible from a single 2 See Index under ravant, PREFACE XIII poem. On account of these self-imposed restrictions Ferdousl's sdhndme became a self-evident choice, and most of the examples come from sdhndme , but for a change the examples used to illustrate the chapter on rhyme are taken from NezamT's eqbalndme*. part two is a description of all but the rarest rhythms occurring in Classical Persian poetry i * * 4 . The first chapter in this part of the book teaches the beginner how to ascertain the rhythm of a given Classical Persian poem. In the next chapter the advanced student is taught how to ascertain, by means of the “circles of the prosodists”, the rhythm of even a fragment o: a poem amounting to only half a line. This use of the “circles of the prosodists" is not my discovery, but I may claim to have rescued it from oblivion, since it has not previously been described in print and is today apparently known only to a very few traditional scholars 5 . The remainder of Part Two describes the several rhythms one by one. The order of presentation is based upon Elwell-Sutton's classification modified here and there for didactic reasons. The rhythms have been illustrated with examples drawn from the whole corpus of Classical Persian poetic literature, including even some lines from recent poems written in the classical tradition. 1 have tried to choose the examples so that they not only illustrate the rhythms in question, but also give the student a glimpse of the beauty and richness of Persian literature. For each rhythm the traditional classification has been given with an account of the curious and intricate i It might surprise the reader that I find it necessary to dwell upon the didactic considerations which led me to choose the examples in Part One from sdhndme, However, I have had occasion to learn that my motives for this choice can be misunderstood. Thus Bo Utas, University of Uppsala, in an evaluation of my manuscript wrote: “The author's basic iramstic training at the University of Tehran often makes itself felt. The native perspective ... seems time and again to lead to a misleading accentuation of the material. This is especially true of the basis for choosing the examples, namely that sdhndme and its metre motaqdreh should have a dominating role in Persian poetry. This can be taken as a reflection in the Tehran University syllabus of the “Great Iranian" ideology of the Pahlavi dynasty," (!) 4 Elwell-Sutton's statistics show 1 that the thirty-three rhythms treated in Part Two cover more than 99% of Classical Persian poetic literature. See Elwell-Sutton, pp. 145-160. 5 The student will probably find it difficult to employ this method since he cannot combine his knowledge of the "circles of the prosodists" with the native's instinctive feci for rhythm, but as l have shown in § 166 he can substitute the principles of ravdnl for his lack of instinct and thus overcome the difficulty. Incidentally this shows that the phonetic qualities of ravdnl are not only of theoretical interest, but also of practical utility. XIV PREFACE mechanisms by which the traditional classifications are arrived at 6 . This part of the work lays no claim to originality. It is intended merely as a convenient presentation of well-know n facts and its indebtedness to Elwcli-Stitton will be clear to any scholar in the field. Then follows a summary description of the works and genres where the rhythm in question is met with, but this is not intended as a history of the genres. After the description of each rhythm there is a list of the poems in that rhythm that occur in the lyrical works of SaSdi, Hafez and Moulavi, and in the great anthology of 2. Safa 7 . part three shows how Classical Persian prosody was adapted to Urdu, Karakhanidic and Ottoman Turkish. The treatment of Urdu is complete. All the differences between the Persian and the Urdu systems are described and illustrated with examples. 1'hese differences were necessitated by corresponding differences in the phonological systems of the two languages. This has not been shown before, and it results in some — as I think — original contributions to our knowledge of the Urdu sound system. Entirely new is my description of the Karakhanidic prosodical system, which has till now defied all attempts at analysis. The chapter on Ottoman prosody is an exposition of the general principles. I would like some day to elaborate my chapter on Ottoman Turkish prosody into a complete description and to add a chapter on Chagatay Turkish prosody. The relationship between Classical Arabic and Classical Persian prosody I have not examined, my knowledge of Arabic being insufficient for this important task. A comparison of the prosodical and phonological systems of the two languages could probably settle the controversy as to whether Persian prosody is derived from Arabic prosody or, as I am inclined to think, it 6 But the terms Xj vateid and < — . sabah, which are considered to be constituent elements of the rokn have not been mentioned, They are useful for understanding Arabic prosody, but to my mind superfluous in a work on Persian prosody. 7 For rare rhythms such lists of references are given in Appendix One. These lists may also be used for a different purpose: When trying to identify a (mono-rhyme) quotation of unknown origin one may first ascertain its rhythm, and then check with the help of these lists of references (and the directions given in §264) whether it occurs in any of the said works. It is true that the references cover only a small part — less than 60,000 verses — of the extant Classical Persian lyric poetry, but it is central and oft-quoted. If other scholars would add similar lists of references covering more Classical Persian authors and anthologies this might in time develop into an important tool for identifying verses of unknown authorship. PREFACE XV was only the Arabic terminology that was adapted to a basically native Persian system. appendix one is a reference list of 217 Classical Persian rhythms and 277 variations on them, appendix two is a reference list of those spellings which cannot be deduced unambiguously from the transcription used. Appendix three is a folding table showing the rhythmical arrangement of MoulavT's divan (as explained in §267). The supplementary notes are amplifications of some paragraphs provoked by criticisms of my manuscript received after the book had gone into press. Finally there is an index by Grethe Thiesen. The transcription of the vowels is my own invention. It may seem a little complicated, as it has to be in a work dealing with both originally long and short and metrically lengthened and shortened vowels. Moreover it also shows the Classical Persian system of phonemes as well as the pronunciation of Classical Persian poetry as it is recited today. For works not dealing with prosody I would recommend the following transcription : a , e, p, e, e>, a, i, i, u, u, ei, ou H . I believe that this transcription of the vowels is an improvement upon the existing systems, though it may at first appear to complicate the represen- tation, but I am not satisfied with my transcription of the consonants which distinguishes phonemes all right, but not graphemes. I chose this simplified system of transcribing the consonants thinking that the juxtaposition of the transcription and the original script would make the distinction of graphemes superfluous in the transcription. Were it not too late to alter now, I should have switched over to the system introduced by Elwell-Sutton in his Persian Metres (with the addition of the sign jr). I have noticed that many advanced students of Persian, though quite fluent in reading printed characters, find it difficult to decipher texts written in the national Persian nasta^llq script. I have therefore had all the Persian examples calligraphed in the national script. Since they are also given in transcription, this should enable the students to accustom themselves to the naslaSliq script without any special effort. Similarly the Urdu and Ottoman Turkish examples have been calligraphed in typical Indian and Turkish calligraphy. During my work with the book I made — sometimes by sheer serendipity — a number of minor discoveries having little or no direct relation to prosody. 8 Cf. 1 3-20. XVI PREJ ACH These have been reported with the utmost brevity in footnotes. I should like to draw attention to footnotes 4 and 5 in § 128, where the date of the Modern Persian vowel shift is fixed 9 . It goes without saying that this work could not have been carried through without help and support from many quarters. My first thanks go to Professor Jes Asmussen, Copenhagen, who not only recommended me for the fellowship during which I wrote the main part of the book and helped me to procure a grant or its publication, but also as my advisor gave me complete freedom to plan and execute the work. I am deeply indebted to my Persian teachers in Tehran University who inspired me with love for Persian culture and most of all to Professor Jamal Rezal, who had the patience to subject my drafts of the first fifteen chapters to a detailed discussion. My friend and fellow-student at Tehran University, J. R.S. Cooper, read through Parts One and Two with me and did his best to change what I had written into English and being himself a Persian scholar made many useful suggestions. Alter returning to Copenhagen I have had recourse to the indefatigable kindness of Dr. Eric Grinstead for correcting my Danish-coloured English. The book has also benefited from the scholarly advice of Professor Hans Hendriksen, Rare Thomsen Hansen, Ahmad TafazzolT and Yann Richard and from the severe, but useful critique of Bo Utas, who wrote an evaluation of the whole work. Also Professor Annemarie Schimmel, Harvard, read through the manuscript and improved upon details of translation and interpretation. Professor D. N. MacKenzie, Gottingen, had the kindness to read the second proofs and point out a number of mistakes. Moreover his keen ear for style has certainly given my book a more scholarly appearance. 1 regret that I did not solicit his help at an earlier stage. 9 Using the information given in footnote 5 the reader wilt also be able to arrive at a more satisfactory interpretation of qazal no. 2542 in Moulavi's divan ( kollidt r farm). It seems to say: O moon, shine on my friend and say: “Friend, eunuch , thou art rottingr O wind, strike his hair (and say): "O beautiful eunuch , thou art rottingr' Whether thou art here or there, whether thou comest or contest not, thou art all sugar and sweetmeat. O wonderful sweetness! Eunuch , thou art rotting? etc. But in fact this qazal should be translated as follows: O moon, shine on my friend and say: "Friend, / love thee?' O wind, strike her hair (and say): “Beauty, / love thee!" etc. PREFACE XVII If the pages of the book are pleasing to the eye the credit goes to the calligraphers. 1 am specially grateful to Professor Alt Alparslan, Istanbul, and K..L. Gupta, Jammu, who offered rec service. For the indispensable task of proofreading I have to thank Bent Huno and Grethe Thicsen. It has been a privilege to cooperate with Dr. Helmut Petzolt, publications manager of Otto Harrassowitz, Wiesbaden, and his staff, who have dealt with accents, quantities and strange scripts in an exemplary partnership with Imprimerie Orientaliste, Leuven. During my stay in Iran the Royal Danish Embassy was always hospitable. Ambassador Troels Munk came to my assistance at a critical period during the Iranian revolution and saw that my books and papers found their way home. Last but not least it gives me great pleasure to record my gratitude to the following Iranian and Indian friends for various combinations of suggestions, criticisms and inspiration: * ^ Jr* ^ Sallerad, 4.7. 1982 Finn Thiesen LIST OF ABBREVIATIONS Cf § / 75 and Appendix One al. — aslam { e ) ^ al-mo *ijam Sams e Qeis: al-mofyam ft ma^ayir e as'iar el-Sajam, ed. Qazvlni, London 1909. as. — aslam (?) b. = bahr (?) B — baddye 9 ES, Elweli-Sutton = L.P. Elweli-Sutton: The Persian Metres , Cambridge 1976. GI, GII, GUI — Z. Safa: ganj e soxan /-///, Tehran 1961. Gibb — E.J. W. Gibb: A History of Ottoman Poetry 1, London 1900. h. — hazaj (?) H divan e hdfez , ed. Qazvlni and Qanl, Tehran 1941. (H ...) ■ — hazalidt kollidt e sams Moulavl: kollidt e sams or divan e kabir, ed. Foruzanfar, Tehran A.H. 1336-1346. Lazar d Gilbert Lazard: Grammaire du person contemporain , Paris 1957. m. — mojtass (?) M — kollidt e sams , see above. mh. — mahzuf Ml molamma $dt mn. = mosamman (?) mq. — - maqsur Mr = mardsi ms. mosaddas (?) ml. = mat vt (ye) nix. = maxbun (?) p.v. = permitted variation (of) XX LIST OF ABBREVIATIONS Q Qq r. Rypka qasd'ed e far si qazalidt e qadim ramal ( e ) Jan Rypka: History of Iranian Literature, Dordrecht 1968 / Jan Rypka: Iranische Li tera turgesch ich te, Leipzig 1959. (For convenience, both English and German translations are cited.) sariS (e) kolliat e sa^dl, ed. Foruql/AStianl, Tehran A.H. 1354. xaflf (e) xavdtim BIBLIOGRAPHY Primary sources. The primary sources for the study of Persian, Urdu and Turkish prosody consist of the whole body of poetic literature extant in these languages, which no single scholar can read in its entirety and which cannot of course be included in the bibliography. The illustrative examples come from the following sources: Re§id Rahmeti Arat: Kutadgu Bitig , Istanbul 1947. Edward G. Browne: A Literary History of Persia I- IV, Cambridge 1956-1959. Edward G. Browne: A Year amongst the Persians, London 1950. L. P. Elwell-Sutton : The Persian Metres, Cambridge 1976. E.J.W. Gibb: A History of Ottoman Poetry VI, London 1909. Hadi Hasan: A Golden Treasury oj Persian Poetry, New Delhi 1972. W.B. Henning Memorial Volume, London 1970. D.J. Matthews and C. Shackle: An Anthology of Classical Urdu Love Lyrics, London 1972. M. L. Rahman : Persian Literature in India during the time of Jahangir and Shah Jahan, Baroda 1970. Muhammad Sadiq: A History of Urdu Literature, London 1964. \rtA-mo oi - oypji oiy-i l \\ro j^aV --- J ijfr JU : JLJl \ ^ £ A «f\ OUT ^ 1 * jl jy ~~ 4i (L,0 1 Yofo j\jf — “ iJ’jfi ®* A i jfji J \ Y* ® V* 0 ^jy ‘-A i ^ ^ 1 ^ ^ ^f£* j\y_ XXI] BIBLIOGRAPHY \\*Y * O' jr ^p f** 0 O'^J ^Ja .5 ^Ia^« O oLiS^ \rn-\no oiy -j+jJ& JJ^ ^ s'w^y^ \TiV 0^- — ^ j ~ jlu-jj : iSj^j^ i£'**~* ^Lp f^Li — i^JUlIIa'S^J eXi ?^>waJ X>«— J j' t£Xc*- ^ V o i 0 \jf- — .^p jiji ^Jp x*j£ ^ xj* I J Li I or \<\V\ ^LUl lH I ,^1 - Y- ^ hj-M OUT oxJ - -\+£ jU-i' jj/U* j : \rt — \rr* jiy - r-\ aUi \roY^rrY ji^* - r-\ o'^j jj oLii : — \r t • 0 '^-* — jlxtl Jlil ^Lp — tjliTlj X-p OLi 5 ~ \ T 1 0 O'jf- — J nrt< > ^ a\^aIj jUaP O'^jO ' Y’ • t - djjj flc*^ — pU*- jf Ol pLj >^V*\ ~ 'y-a> — t, I Ip o'^i \^V\ — (_jv* ^ \ a^U-aLj< \^V^ a^JU- — ^-Sx- o O-x : -Uj-I ^yLi mv-mi J^ixi - -/y*. - f f oTy \WA *£Jp - dfc ^ A * * AX t^pUt jijl ^Uxl : ol*>- ^*J \ V 1 1 [o'^ff] ^Jp *1*--*^ *> 1 sjz UVjkA o'_^j jA\jJ^ OLa^J' Q - \f$A oV “ ~*jL oii : U^ a .* yfcllx* \ *“\ JA*S"" 0'_JjJ 1 j i W-> US'" I BIBLIOGRAPHY XXIII \rtv\rrv jiy 0>. <_r-^ j ji-OI J^U- ^UT J yrvt-yrvy jiy - UAl - jU^I ^ j r Uj^l r ^b - T-y jy. our y To t 01^- — aU' f l^L — j \ j * I : j y y*®yr oi^ j ui y ^ t V (^juLb pl^b — a^U Li' ; ^Ual WIN j5"b — ~ A^biyi ; — y^V jfl - a^ljJp ft*b * jl^-Vl jjit : — yry\ oiy - JV. ur> >-** j JIH : ^ -u. yr« t uiy - oipt^ j j>i : Ju» jjoii J^u Qaem-Maqam e Tupci’s rofoKi” was copied from a tablet on display in the Museum of Fin e Kasan. In the Appendix to Part One the exact provenance is given for all passages quoted in this part of the book, In Parts Two and Three where textual accuracy is not so much at stake, this procedure has not been deemed necessary. Secondary sources. This list confines itself to essentials and should be supplemented with the information given in the footnotes and the standard bibliographies (see Rypka 757-759/572-574). Walter G. Andrews: An Introduction to Ottoman Poetry , Minneapolis 1976. H. Blochmann: The Prosody of the Persians , Calcutta 1872. Edward G. Browne: A Literary History of Persia II, Cambridge 1956, L. P, Elwel I -Sutton : Persian Metres , Cambridge 1976. Hermann Ethe: Neupersische Litteratur, in Geiger / Kuhn: Grundriss der iranischen Philologie II, Strassburg 1896-1904. Massuud Farzaad: The Metre of the Robdaii, Tehran 1942. Mass’ uud Farzaad: Concise Persian Prosody, Coventry 1966. Mass uud Farzaad: Persian Poetic Metres, Leiden 1967. XXIV BIBLIOGRAPHY J. Garcin de Tassy: Rheiorique et prosodic ties longues de l orient musulman, 1873. E. J. W. Gibb: A History of Ottoman Poetry /, London 1900. Francis Gladwin: Dissertations on the Rhetoric, Prosody and Rhyme of the Persians, London 1801. K. M. Jhaveri: Outline of Persian Prosody with Figures of Speech, Bombay 1922. F. Meter: Die schone Mahsati, Wiesbaden 1963. G. D. Pybus : A Text Book of Urdu Prosody and Rhetorics , Lahore 1924. H. Ritter: Die Geheimnisse der l Vortkunst des 9 Ahdaiqdhir aTCurcanJ, Wiesbaden 1959. Friedrich Riickert: Gramma tik, Poetik und Rhetorik der Perser , Gotha 1874. Ralph Russell: A Primer of Urdu Verse Metre ; London 1974. Jan Rypka: La metrique du mutaqarih epique person , Travaux du Cercle Linguistique de Prague 6 (1936). B. 1. Sirus: Aruzi tojikl Dushanbe 1963. Ahmet Talat: Turk siirlerinin vezni , Istanbul 1933. Felix Tauer: Persian Learned Literature from its Beginnings up to the End of the J8th Century, in Jan Rypka: History of Iranian Literature , Dordrecht 1968, - J J - 5 jS** \^\0 jj. JlT- l 1 J * \nn oi - J'yui \ ^ 0 1 J Im# I j m b 4-P^U t j+m \ 1 L^J I *A*S' ~ a*3S^ •- A iJ • Jr-'*" - u ^' \ TT \ ol>- BIBLIOGRAPHY XXV 1 A*W OUT - ^j> 3j e. T ^ X* o o \j^ A3 Li j * [^m] o\jf* iSy^j j^* \rol U'V “ : — \fto jl^ — ^ijj ~ OU-Vl jj£ : c^j'^w aJ>0' c~*&j \fYi ^jJ' ^.L?i *j ^VY* u ' jr — jUOV' Jl*a : j^jJ' ^r~0 J — *" ~ u^Jj* ‘ v_?^'—1 j^' \rrv oty - Ju> <3 'TM * r^ —> “ c/>* crAJj* : Ji-^ 1 ^ J* \rto J'^r — a* If l£^o : 3 U/ ->_>■*—-* \ A > t a^J 5" - OP%]l jjlJb- : ^ j,0)1 J (►***! 'j fc r^' 3 ^3' ‘j'jr* - JL^y^ j£jf j> 3 ST * A JJa.J - ^Y*yv o'^ ^•'jU <3r^ • y.iy. STTS JV - ^-jli * * ^jj : ^r*Y Ob — ,17 u'O „0l^l ^3 ^ T S £ aOxJo^j — 1 olplw? j C-P^l djj : jU® Jjjjl J^U- PART ONE THE SYLLABLE “There neither is nor can be any essential difference between the language of prose and metrical composition". William Wordsworth I THE VOWEL SYSTEMS OF CLASSICAL AND MODERN PERSIAN § 1 The basis of Classical Persian prosody is the division of syllables into short and long (and overlong) syllables. In order to perceive the rhythm and appreciate the different metres one must therefore be able to distinguish between long and short syllables. Let us begin by considering the vowel system of Classical Persian. § 2 Classical Persian had three short vowels : a , i\ w, five long vowels : a, u , e , o, and two diphthongs: ai and au. By native grammarians / and u are termed #1 ya' e maSruf and j\j iw e ma Sruf respectively; f and o arc called #1 rtf' e majhul and jb i'dr e majhul. i and « are called maSruf "known” because they occur in Arabic and thus are known to the Arabs, e and <5 do not occur in Arabic; hence they are majhul "unknown” (to the Arabs). The so-called "long diphthongs'' at. ui etc. are better treated as combinations of long vowels plus consonant: dr. fir etc. §3 In Modern Persian a distinction of vowel quality has taken the place of the Classical Persian distinction of vowel quantity and the number of vowel phonemes has been reduced as shown below: Classical Persian a > > Modern Persian a [but final a becomes e\ see § 16] e [but before a vowel / remains; sec §93] i 4 THF SYLLABLE §3 u a l e u 6 ai au > 0 > a 1 > 1 > 1 ■a 1 A A u > ei > ou Examples: jLl 1ST guiistdn > go/estdn jlu-y bo stan > bus tan iskandarndma > eskandarname jl yl eran > iran Modern Persian sir may be either from ser Similarly, yal dhu may be either from aho "fault” “lion” or sir "milk”, or dhu “gazelle”. §4 The Classical Persian pronunciation is shown in F. Steingass: A Comprehensive Persian* English Dictionary, London 1892 (several impressions, still available). Note that for e and 6 he writes e and o without macron. Thus, when he writes ro:i umed u him, we would pronounce it ru: e omid o him j -V 1 “The day of hope and fear i.e. "The Day of Judgement”). Unfortunately Steingass is not too reliable in his transcriptions. More reliable information may be found in Paul Horn: Grundriss der neupersischen Etymologic, Strassburg 1893 (reprinted in Hildesheim 1974). and in Fritz WolfT: Glassar :u Firdosis Schahname, Berlin 1935 (reprinted 1965). The Modern Persian pronunciation is shown in Junker/Alavi: Persisch-deutsches Worterbuch, Leipzig 1965, in A.K.S. Lambton: Persian Focahu/ary. Cambridge 1954, etc. Curiously enough many Persian scholars have fallen, so to speak, between two stools, and use a system of transcription, which represents the Classical Persian pronunciation, but leaves out d and e (writing « and /), thus making the number of symbols agree with the phonemic structure of Modern Persian (thus E.G. Browne in his Literary History of Persia). This is unfortunate because in the case of the older poets a correct distinction between i, u and e . a may help in detecting spurious verses and in correct interpretation. Of. §131. §5 Indo-Persian has preserved the Classical Persian pronunciation of the vowels. Unfortunately, due to the cultural co-operation between Iran and India (and Pakistan) every year a number of young teachers of Persian return home from Iran to teach their compatriots the "correct'' Tehrani pronunciation, so that m another generation or two the Classical Persian pronunciation will have been forgotten. 1 It is unnecessary here to take changes like an > un into consideration. THE PERSIAN VOWEL SYSTEMS 5 §9 In the Persian of Afghanistan, too. mujhul 6 and e arc still distinguished from maSruf u and /. The Afghan pronunciation is shown in L.N, Kiseleva and V.I. Mikolayttk's Dari-Russkiy Slovar', Moskva 1978. §6 Although, as we have seen, a distinction of vowel quality has succeeded the Classical Persian distinction of quantity, still, the Persians are taught that a, e and o are kutah “short”, whereas d, / etc. are xJLi holand "long**. Thus the a in ^ sabr “patience” is called “short”, though in fact it is pronounced considerably longer than i in J mikonam "I do”, which is in turn called “long”. The result is ihai even very competent native scholars often have the strangest notions about long and short vowels. Thus our teacher of prosody at the University of Tehran once asked my help in ascertaining ihc metrical set -up of a poem by Victor Hugo, which he had found impossible to scan. It turned out that he had a prion taken alt the e's and o's in Trench to be reckoned short and the i's and u's to be reckoned long, whereas he did not know what to do with the y's* the eau\ etc. §7 However even in Modern Persian there are still a number of features which distinguish the originally long vowels from the originally short vowels; they have as we shall see preserved some of the characteristics of the Classical Persian long vowels. The reader is referred to Gilbert La/ard : Crammaire tin per sun contemporain ’ §7. where the author proposes the terms stable vowels Tor the originally long vowels and unstable vowels for the originally shore vowels, §8 As a result of the phonetic changes which have taken place since the canons became fixed, ability to read and recite Persian poetry rhythmically is not something the native Persian is born with. Contrary to what many foreign scholars believe, it requires no mean degree of education and practice. A good reciter of poetry carefully pronounces d, /, and u long whereas a , e , and o are pronounced short. The diphthongs ei and ou are of course pronounced long. §9 Although a significant amount of free verse ( jljT yul .f ri oflddun "to fall", the reason for this is not that his "poetic licence" allows him to change the words at will whenever the rhythm requires it — he would never say *atd instead of t id "till" nor * us tor instead of /-Z 1 os tar "camel" — aha and ufiddan are simply older forms of the words, which have been preserved in the poetic language, cf. Pahlavi abtig and oftadon*. §12 However, most works on prosody do not devote many pages to these aspects. Instead, they preoccupy themselves with the barren and pedantic discipline of giving a different name or number to each possible rhythmic variety. The latest work in the field of Persian prosody. 3 Pahlavi words are everywhere quoted according to D.N. Mackenzie: A Concise Pahlavi Dictionary , London (O.U.P.) I97L Reprinted in Tehran at EnteSarat e Natnard. THE PERSIAN VOWEL SYSTEMS 7 $12 L P. F.lwell-Suttoti: Persian Metres*, is no exception (cf. my review in Acta Orient alia jV, pp. 242-257. Copenliagen |97X)-'. A huge anti complicated terminology has been evolved by the traditional Arabic and Persian prosodists. In this hook only the terms commonly met with have been defined. However, in Appendix One the traditional designations of all metres met with in Classical Persian literature have been given. For further information on the traditional terminology readers may refer to Elwcll-Suilon. 4 Cambridge 1976. This work is hereafter referred to as Elwell-Sutton. 5 I now think that the criticism in my review- is too severe. Though the censures arc perhaps justified, yet I feel that I should have dwelt more upon the merits of the book, which are the outcome of original research and patient labour. The present Introduction does not make Eiwcll-Sutton’s work superfluous and the advanced student will find Persian Metres very useful. [[ TRANSCRIPTION §13 As mentioned in §8, good reciters of poetry carefully pronounce a, /, and u long, whereas a, e , and o are pronounced short. In accordance with this we transcribe as follows: Vowels prosodically short, i.e. those which are generally not shown in the Arabic script, are represented by a, e, and o. Vowels prosodically long, which are generally indicated in the Arabic script by the letters !, l£, J, are represented by d f /, and u. Strictly speaking we should transcribe a instead of a , but for typographical considerations the latter has been preferred. §14 Quite often the metre requires that e and o should be pronounced long. They are then transcribed e and 5. Similarly i and u are not infrequently to be pronounced short. They are then transcribed by / and u. §15 More rarely, the metre requires a to be pronounced long; it is then transcribed a:. Very rarely a must be pronounced short; it is then tran- scribed d. The purpose of this little inconsistency of transcription is to keep apart the distinct phonemes a and a [&). §16 Classical Persian a often becomes Modern Persian e notably when final, and in the preposition t he < ha "to”, but also in other cases 1 . It is difficult to decide how far one should follow the popular pronunciation in this respect. Shall we write moStamed or mo^tamad for “trustworthy"? Shall wc transcribe 4 ^-U* “(medical) treatment" by moSdleje. moSdlaje. or mo^diajal I may have gone rather far in substituting e for a. To make up for this, e has been differentiated by a diacritical point whenever it corresponds to Classical Persian n, thus mo blamed, morale}?. § 17 When following an originally labiovelar fricative, Classical Persian a has become Modern 1 For these vowel changes refer to Lazard § 10 and §20. §20 TRANSCRIPTION 9 Persian (k thus v h f tnl > xord 'he ale"* This o too has been given a diacritical point: v »^rcA (But when required hy the rhyme the original ti has been allowed to stand). § 18 Likewi sc / and u have been marked with diacritical point whenever (hey correspond to the Classical Persian tnajhu! vowels e and d. §19 It would have been quite logical to transcribe similarly the diphthongs by ci and ou * bm this has been considered unnecessary, and they are transcribed simply W and ou. §20 Transcription of the vowels: a a: e c e i e o 5 0 1 a a i ■ i i ■ i m- u u u * u ■ ei ou — Classical Persian pronunciation a Modern Persian pronunciation a the same metrically lengthened = the same metrically lengthened ™ . a e - the same metrically lengthened = ... u o = the same metrically lengthened a o a l... the same metrically shortened lsv the ci- ■ the J* = y the same metrically shortened e i same metrically shortened u u same metrically shortened 6 u the same metrically shortened l£... - J at au nim-fathe , see Compare the following transcriptions: behest "paradise £ geristan "to weep" budan "to be” I*' ei ou -38. Ui bejd "proper * i gorixtan “to flee” anduxtan “to collect” 10 THE SYLLABLE §21 As all examples are given both in transcription and in the original script, 1 have not thought it necessary to distinguish in the transcription between the various Arabic s's and z’s etc. A glance at the version in the original script will immediately show whether e.g. r is intended for r, r, or r. Technical terms, names of authors, etc., are given in the original script only at their first occurrence, but for the convenience of readers with short memory their original spellings are repeated in Appendix Two. Apart from this, little needs to he said about the transcription of the consonants. I have in the main followed the system used by A.K.S, Lamhton in her Persian Grammar. Cambridge 1953. but I have not distinguished d from £_ both being represented by be it (plural OLI ahydt). The heir is the basic unit in Persian verse and it always consists of two rhythmically identical (or near identical) halves known as f' 1 j-a* mesrdS (plural Ai j La* masari*!). These again are divided up into a number of feetj^j rokn (fmiral jlSjl arkan). §25 The traditional terminology of Persian prosody !*• Seim e Saruz) has been borrowed from the Arabs. By a ralher far-fetched analogy most of the terms arc named after parts of the tent. Thus heir means "house' 1 (i.e. "tent "); each of the two Haps of the tent-door is called ntesra S, and rokn means "pillar" (i.e. "the prop of the tent"). §26 hcii and mesra S are often translated "distich" and "hemistich", but these translations are not satisfactory 3 , so we shall in the following simply use the traditional terms heir. w.vrcJV, and rokn. (Note the plurals of heit and rokn : ubyut and arkan ') In Persian the Arabic form mesra S is often shortened to mesraS. and the term nim-heii “half-Ae//" is also used. 1 Pronounce: tlo-ru-qi. 2 Cf. E G. Browne: Literary History of Persia ii. Cambridge 1956, pp. 24-25. 12 THE SYLLABLE §27 §27 [n the above example (§23) each mesra ¥ contains four arkan 3 , and each rokn consists of one short and two long syllables: except the last one which is apocopated: thus: be it mesra mesra S rokn rokn rokn rokn rokn rokn rokn rokn This is probably the rhythm ( J jj vazn) most frequently met with in Persian poetry, and Ferdousfs sahndme and SaSdi's bustan follow this rhythm. It is known as j£- bahr e moiaqareb “the nearing metre". Since each beit comprises eight arkan it is called mosamman “eight-fold" (i.e. octameter). Then it is apocopated, in the above instance termed mahzuf. In short, it is the bahr e moiaqareb e mosamman e mahzuf 4 , but as this is the commonest variety of bahr e moiaqareb, it is generally referred to simply as bahr e moiaqareb. Up to § 147 all examples 5 will be in this metre, so the reader is advised to note carefully its rhythmical structure (vazn). §28 The rokn can be defined as a recurring section of a rhythmical pattern. According to Elwell-Sutton (p 85) the subdivision of a mesra*! into arkan is “not ... very profitable ... and may even be misleading. In a purely quantitative metre ... the division of a line into ‘feet’ cannot be more than a convenience, a mnemonic device ". However, in the case of a rhythm like w w - „ - - „ it appears to me perfectly natural and justified to subdivide it into arkan m (he way it has been done by the prosodists. Moreover, there is also a linguistic argument tn favour of this subdivision. The last syllable of a word more often coincides with the last syllable of a rokn than with other syllables of the rokn. Thus in the first chapter of sdfmante 65.5" u of (he final syllables of the arkan coincide with the final syllable of a word, whereas for the other syllables of the arkan the percentage is only 48"„, (In this analysis the last syllable of the mesra k was, of course, not taken into account). Where internal rhyme occurs, this too seems to favour the subdivision of the mesra V into arkan, cf. §§ 137. 149, and 192. In certain metres, however, the traditional subdivision of the mesra*! into arkan is indeed artificial and misleading. This problem will be dealt with in Part Two. 3 Plural of rokn. 4 Cf. Chapter XVII. for ha/ir see especially § 176. 5 With seven exceptions to be found in §§71, 82, 100, 111, 112, 126 and 127. IV OPEN AND CLOSED SYLLABLES §29 In the beit quoted in §23 the rhythmical effect was produced exclusively by the contrast of long and short vowels. This rhythmical effect is, however, as frequently (if not more frequently) brought about by the contrast of open and closed syllables. An open syllable consists of consonant + vowel (e.g. j ze “from”); a closed syllable consists of consonant + vowel + consonant (e.g. man “I”). §30 It might be objected that some syllables do begin with a vowel. However, according to the native Perstan grammarians, even words like ^->1 dh '‘water" and jj jA emrur "today" begin with a consonant, namely hamze "glottal stop", thus ah, 'emruz. As far as Persian is concerned this view is indeed open to doubt, but in a treatise on prosody it may be conveniently accepted. §31 Now consider the following mesra ‘j ; •» v • J I • ** M * *# tahamtan nayamad be laskar ze dost “Tahamtan (i.e. Rostam) did not come to the camp from the plain” l . Here only one of the long syllables contains a long vowel: td; elsewhere the length-effect is provided for by the consonants following the short vowel: ham , tan, mad , las, etc. §32 When the Persians want to describe the rhythm of a particular rokn , they do so by quoting a derivative of the Arab root . Ui faSala, which rhythmically corresponds to that rokn. Thus tahamtan is said to be Ojj ^ bar vazn e fa^ulon “in the rhythm of fa^ulon" , and the whole mesra V is bar vazn e fa^ulon jd^ulon fa^uhn JaSul. I This nwsrdS is taken from saJtndnw. Henceforth, when nothing else is indicated the quotations are from xahnamf. 14 THE SYLLABLE The rhythm of each mesra V of the heir quoted in §23 would have been described as fa Sulim faSiiltm taSaL The difference between J yk faSul and U# faSal will be explained in §42. Note that faSuhm is written J y* as a term of grammar, but y* as a term of prosody. §33 Another less commonly used way of indicating the rhythm is to use an appropriate combination of the syllables ia and tan. Th or Jw ta-hm-tan = faSuhn. §34 The traditional prosodisls also had a method of representing the rhythm graphically* the so-called — ** hesah c V aru : 'The reckoning of prosody ". A short syllable is represented by a and a long one by \o. Thus « would be rendered \a\oa< (Read from right to left!) o is said to represent a consonant followed by a (short) vowel ( ^ harfe motaharrek ), whereas represents a consonant not followed by any vowel ( hurl v sakvn). Thus long syllables - w hether made up by the combination consonant -t vonef + consonant or by the combination const* turn! 4 tong yowcf — muy be shown by Uh since the long vowels are thought to be- as in the Arabic script indeed they are -combinations of the short vowels. a > t\ (K and the consonants, \ v\ w This concept of the long vowels at once becomes comprehensible when we consider either the Arab or the Classical Persian pronunciation. Then a equals a\ i equals i\\ and u equals uv. According to this system the mesra ? quoted in §31 should be represented as OO OjfHP P OO O OO. Since the last rokn in the mescal — ze itast — ends with a double consonant, it is rendered 'oo instead of 00 . V OVERLONG SYLLABLES §35 A very special feature of Classical Persian prosody is its distinction between JLL boiand “long" and j ' derdz “overlong” syllables. An overlong syllable is equal to a long syllable + any consonant (except j th cf. §§83-88). Thus j* sar “head" is long, but j y sard “cold" is overlong; U mu “we" is long, but jU mar “snake” is overlong, y mb “hair” is long; the doublet iSj* mbv is overlong. §36 Except at the end of a mesrd V (cf. §41). an overlong syllable is reckoned equal to the combination of one long and one short syllable. sard and mar accordingly count as - w. In the transcription we indicate this by a raised \ thus sard*, mar*. Consider the following four examples: , ,j£ A /- | manam tan- g'del fa sodam tan- g'dast I am sad since I have become poor”, As j c ; l » H ** be mind 1 magar hi - namai bci - z* sad “Perhaps in Heaven I shall happily see you again”. cera d- madi ba- z* pisam begu “Why have you come back to me? say”! Or mum. 16 THE SYLLABLE X '/> '/X 1 At 1 ttir /m £tf,v- / ’ badx"a - 1 ft* baxt "The ill-wishing fate was imminent ". §37 In traditional Turkish and Indian recitation of Persian poetry, the ’ of the overlong syllable is realized as an indistinct vowel, known as nim-fathe. In Persia proper this 'is not pronounced, the very length of the overlong syllables, even without \ being sufficient to convey the rhythm. A glance at the representation below of the four examples quoted in the previous paragraph, will immediately show how the rhythm persists although ’ is not pronounced. (The overlong syllables are here represented simply by "overlong strokes"): VX “ — v “ “ ix — W — — W A sonograph analysis of recordings of recitations of Classical Persian poetry made by Elwell-Suiton (pp, 193-207) shows an average length of 0,21 see. for short syllables. 0,33 see. for long syllables and 0.59 sec. for overlong syllables. §38 Whether this 9 was ever pronounced in Persia proper is a moot point. It may never have been pronounced at all. However, the following facts seem to indicate that ^ was originally pronounced: a) Persian is not fond of consonant dusters. The appearance of an anaptyctic vowel after overlong syllables is therefore phonetically plausible. b) The traditional Indian pronunciation of Persian has preserved many archaic features (cf. §5f The I ndo- Persian tum-faihf may well be another such archaic feature. c) Persian loanwords in Hindi and Turkish often seem to preserve the nhn-fathe: Turkish Kamuran (n.p) from Persian kantrmi fkdmran). Hindi kdrigur "artisan ' from Persian kargar (kdr'gar). d) F.ven many Modem Persian words seem to preserve the *: mehraban - mrltr-hwi "kind", similarly nurtlehun or nardohun "staircase d e) In the colloquial language this apparent preservation of * is even more common (d Lazard §18), A most illuminating example -not mentioned by Lazard is ij* j* P* r muni "an old man ' as compared with jamn manl "a generous youth '. In ordinary speech 2 The Classical Persian la bio* velar fricative phoneme x* (written ) is always reckoned as a single consonant. In Modern Persian ir has become velar and is not distinguished from x (r-) in pronunciation. §40 OVERLONG SYLLABLES 17 the former is mostiy pronounced pir e murd with un irregular ezdle 3 , whereas the latter is never pronounced * Javan e mard, this being in perfect agreement with the prosodisis' rules for the pronunciation of ntnofdthe (cf. above and $$83-88). Also d/arin “bravo" as compared to jtyv nejrin "fie, hatred" shows agreement with the prosodisis rules. Other examples are Jl^T dsmdn "heaven" and sabr kon "wail", often pronounced axeman and sabre kon (or sah kon). The truth may be that ■' was pronounced in some dialects and not in others. At any rate ■’ is not pronounced in present-day Persian recitation and in accordance with this a raised letter has been chosen for the transcription, the principle being ihui raised letters indicate sounds not to be pronounced. §39 When an overlong syllable is followed by yet another unvowelled consonant as in 'gust “meat Jjl5" kind" knife", fdrs “Persis”, 1 kask “would that 1 ', this last consonant is not reckoned. Such consonants we indicate by raised letters us in the following mesra V: II tl j if i / hand x'd- s 1 -' kardan “He only wanted to find you”. lord x"d- s'tdr Here the / in x'dst is not reckoned. See also the third example in § and the example in §233. In very rare cases the second consonant in a duster following a short vowel is not reckoned. This is noi considered quite correct. Lxamples hereof may be seen in Myulavi's divan, the so-called kotlidt e sains or divan e ktihir 4 . e.g. line no. 34666 where pan/samhe "Thursday" must be scanned pahsamb^. line no. 13089 where soda ram/ "Master" must be scanned xodiivan, and line no. 21970 where an ast must be scanned an as. §40 While reciting such verses the Persians sometimes leave out the extra consonant, but more often they do pronounce it It ns noteworthy that in the colloquial language, loo, a consonant following an overlong syllable is often not pronounced: ^.LkT kojast becomes kojas “where is”; — Jj rasigu becomes ra.sgu "telling the truth” (cf hazard § 17,3), The rule stated in the previous paragraph shows that this feature of colloquial pronunciation is probably a thousand years old. This is corroborated by the fact that the word c— C rast "straight, true” has been borrowed by the Turks In the form rust as well as ms. 3 For a different explanation see L. P, Elwdl-Sulton: Elementary Persian Grammar (Cambridge 1969), p, 42. Cf Additional Notes below. 4 Edited in ten volumes by Foriizanfar, Teheran A.H, 1336-1346. Hereafter referred to as koUiat c settns* 18 THE SYLLABLE §41 The last syllable in a mvsraS is anceps , i.c. indifferent with regard to length ; it may be short, long or overlong. Anceps finals are well known in the prosodies of other languages, e g. Greek and Sanskrit, and the explanation is simple: since the final syllable is followed by a pause, it does not change the general rhythmical effect, whether it is long or short. When representing graphically a particular rhythm, the final syllable will for convenience always be represented by whatever its factual length may be. Cf. § 166 end. §42 Tiic traditional prosodies, however, distinguish between long and overlong finals. This distinction serves no practical purpose, hut does make the already prolix terminology even more complicated. Thus if the final rokit in the metre with which we are dealing ends with a long syllable (foo or faStth cl> §§32 and 34) it is said to he mahzuf "cut off*, but if it ends with an overtong syllable TV oo or cf, §532 and 34) ii is said to be maqmr "shortened'’. The metre of sahndme must then he described as fuihr v nmUufarvh t* nittsaninutn t* mahzuf According to Elwell-Sulton short e and o are approximately one and a half limes as common as long e and d in Ferdousi and MoulavI t whereas in Hafez short e is twice as common as long e, and short o five times as common as long o n , My own counts yield the following results for Hafez: In 495 uhyai a total number of 842 short t\ 359 long e, 310 short o , 68 long 6. It is clear from Hafez that the long e is less stmgm "heavy" than the long d. Cf. §§ 50 and 76. §58 Prosodically long ezdfe (e) is sometimes indicated by a vertical kasre, known as JU kasrf ye monuil (,..), thus in Nicholson's edition of mas navi re nwHnavi. In most texts, of course, ezdfe whether long or short is normally not shown at alt §59 In poetry j to be pronounced va occurs only at the beginning of a mesra c i and even in this position va is extremely rare. . A 4 .. V tyiSjjU (j i * X y yU negahbd- n e ganji to az dos- mandn va dunes negahbd- n e to jd- veddn “You are the guardian of a treasure against enemies, and knowledge is your guardian all the time”. y Bu Sakur cf. § 181. The language of poetry here agrees with the colloquial language, not with literary prose, where va (which is an Arabic loanword) is quite common. Some native scholars are of the opinion that j should never be pronounced va in poetry: at the beginning of a mesra C it should be pronounced vo or o . 11 Elwdl-Sutton p. 213. Cf also §51 above (with footnotes). 26 THE SYLLABLE §59 Another example of va may be seen in §65 (second heit). Cf. also §54 (foolnoie 3) and §260 (footnote 15). §60 When followed by a vowel j may — if the metre requires — be pronounced v. This v may occur before any word beginning with a vowel, but it is especially common in the combinations jl j v az “and from”, jTI j V agar or jl j v ar "and if", aJ S\ j v agar na or aJ j\ j v ar na "and if not", "else", "otherwise". These combinations are generally conceived and written as single words jj va:, vagar l2 ,jj var, aJ^Sj vagarna 1 2 , Ajjj varna 1 3 . /,v, -smZ j i/j g * * be ganj o fozuni nagiri ferib be pis ar faraz a- yadat v ar nesih “Don’t get deceived by riches and greatness, whether you are on your way up or down T ' t J) neha "*' r v a borun 1 5 nomun 1 ~" 14 " - s ’kara darun 6 xerad ra be darga- h e u rah- “The hidden and the manifest, the interior and the exterior, to the discerning mind (it is all) a guide to His court". Nezami J zartr e # co jdmd- gera'md- s pJ 1 9 dastu- ye r e ndba- k'dar 12 Some analyse vagar and vagarna as va-gar(na). However the facts stated in §59 as well as the comparison with var and varna (which cannot possibly be analysed va-rfna). but must be v-ar(na)) make the analysis v~agar(na) much more probable. 13 varna has been borrowed into Hindi ( varna "else, otherwise”, cf §67). 14 For the treatment of n see §83. 15 Cf. § 118 end. 16 Cf §118 footnote 13. 17 For the treatment of n see §83. IK Cf. §§92 and 93. 19 Cf. §39. §61 KASR£ YE EZAFE AND VAV E UTF 27 "The noble Zarir and Esfandiar, as Cf. §§ 79-8 1 . as the intrepid vizier, Jamasp". §61 When the students in Persian schools scan poetry on ihc blackboard, they write below the heii to be scanned a line called the Jiyii- malfuz "pronounced”, in which they indicate — as Tar as the inadequate Arabic script allows— how the heit in question should be scanned. Here e and & arc written ^ and } respectively, whereas short o is indicated with aw * zamme (*?.). The nhn-farhe* too, is indicated by zamme. Hence, while scanning poetry, the Persian pronounces e and d as i and u. and * becomes u. but while reciting poetry l his pronunciation is never heard. The following heii . * for u raj- t o bar raf - r *• ruz e nahard be mahi man e am* n o bar ma h* gard "Down went and up went on the day of battle: to the fish 20 the moisture of blood, and to the moon the dust”. in the malfuz "pronounced" version becomes r 1 ruz e nahard man e am- # ## V forumj tohurruf toruzi cr §^85, 91 and 151 (end). nahard hen ut hi mum. xu no hu raid hogurd 20 On which the earth is supposed to rest. VIII FINAL SHORT VOWELS §62 A final short vowel indicated in the Arabic script by a (-e, -e, -a) or by 3 (’ 0 ) may be pronounced long (-<% -e, -a;', - o ) when the metre requires. The traditional prosodists explain this by a reverse rule stating that final he and vov in such words may be elided when the metre so requires. §63 i he lengthening of e ( kasre ye ezdfe) and o (iwv e dtf ) mentioned in §54 may be considered a special instance of the above rule. §64 Quite frequent is the lengthening of -c\ as in the following example where it is twice to be read long ( hame and gerefte ) and once short (baste): SjSs • • hame rih \ e dhan gerefte be zar derajs e siah 2 has - te bar xii- d* bar “The iron all gilt, a black sash wound round the helmet”. §65 -e and -a — as in ke ”who", “that", wi na “no”, “not”— are rarely lengthened. a?- re “what”, ¥ se "three yek i pi- I’tan di- dam 6 si - r ’cang na has 6 na ddnes na rdv 3 d na sang “I saw an elephant-bodied and lion-clawed (man, hence 1 had no longer) neither judgement nor knowledge, neither intelligence nor dignity”. Here we see no fewer than four occurrences of short -a in the second mesrdk. In the following example -e is everywhere short: 1 cr. §i5. 2 Cf §§92 and 93 for the short i 3 Cf. §54 footnote 10. §65 PINAL SHORT VOWtLS 29 va* digar* ke giti naddrad derang saruy e sepanjt ce pahn 6 ce tang “Moreover the world has no permanency. The house of three and five (o: the transitory world)'’, (what does it matter) whether (it be) broad or narrow (i.e. whether you are rich or poor)?" The rare -e and -a: are seen in the following examples: nebestan be xosrou * bidmu- x’tand delas rd be ddnes har ajr ti- x’tand nebestan yek i na: * * ke nazdi- k e St ce runv ce tdzi yo H ce pa- r’Si “They taught the king to write, illuminated his heart with knowledge; (they taught him) not to write one, but close to thirty (scripts), whether Greek, Arabic or Persian". A # *• #* ze pisas beraftan- d' har se be ham sode sar por az ki- n o del fid dezam “Together all three went away from him with (lit. having become) the head full of rancour and the hearts furious" 4 5 6 7 8 9 . 4 Cf. §59. 5 digar goes back to Pahiavi dudigar and j and 3 are easily confused It is therefore possible that va digar here and elsewhere should be emended to dodigar. 6 This is how this expression is understood in Iran today, However, Pahiavi aspitij means ‘'hospitality", so the sense of sardy c Sepanjt must originally have been “guesthouse" (o: transitory world). 7 For beydmuxnand. Cf. §93. 8 Cf. §55 end. 9 In sdhname this heir is the continuation of the last beit in §60. 30 THE SYLLABLE §66 §66 The ending -e goes back to Pahlavi -ag (preserved in plurals like haveegdn “children ', etc f According to some 10 it was pronounced -ah at the time of Ferdousi This cannot he true; such a pronunciation would make the national epic of the Persians unrhythmical. In the poetical language of the time, however, an archaic pronunciation -ah might have coexisted with the ordinary pronunciation -a. In this case the forms gerefk and hast? in §64 must have been pronounced giriftah and haste at the time of Ferdousi The above explanation is not impossible, but it is far more likely that -ah and -a never coexisted and that right from the beginning final short vowels were capable of metrical lengthening. Lengthening of final short vowels is well know n in Avestan ll , In Old Persian, too, final short vowels are invariably written long, hut the phonetical interpretation of this is doubtful §67 As we have seen - e (Classical Persian **/) is quite frequently lengthened, whereas (Classical Persian -/) is rarely lengthened In this connection it is noteworthy, that corresponds to -a (long!), but -e to -/ (short!) in Persian loanwords in Hindi, Thus mw hatve "child becomes ha via, whereas u hi ke "so that” becomes taki iy -e ‘V §68 -a occurs only in the word na "no", **not'\ which as we have seen is rarely lengthened to mi:. The occurrence of na: is made even more rare by the fact that when a tong syllable is required many poets prefer the synonym j ni (< Pahlavi ne). Thus in a description of the soul o-c- 1 Asadi says: & J V ; ** H * na drd * w>/i/r & na az jd - v* hjru- “(lt is) neither seeking rest nor accepting movement, neither outside space nor occupying space ". na jomhes- pazir n o ni jd* \ k ‘'gh §69 As a prefix U na is substituted for ; Thus, describing the prophet* Sa c di says: $ vat ini i ke ndkar- m E katobxd- tie ve can- ■ * na- whenever the metre requires a long vowel. Srf C/5? «/ mm de ifor d**'* dorost mei/at hesost 10 See e g. H.W\ Bailey's article The Persian Language, p. 177 in The Legacy of Persia. Oxford 1953. If See A.V, Williams Jackson: An Avesta Grammar §§24-26, Stuttgart 1892, 12 See R G. Kent: Old Persian, §36, New Haven, Connecticut 1953. 13 Dr, Jamal Rezat informs me, that the Xordsdni dialects show a similar development of -e and -e> 14 For - d » see §83. FINAL SHORT VOWLLS 31 §71 "(He was) an orphan who — noi having made perfect the reading ( y. not havmg learned to read. being himself illiterate) washed away the literature of several nations". §70 -e is nol lengthened before enclitics like the indefinite article i, the enclitic forms of hudan: - am , -/, ast, -ini, -id, -and, the enclitic forms of the pronouns: -am, -at, -as, -(e) man, -(e) tan, -(e)san, and ye! ye (ezafe). The reason for this is, of course, that in this case -e is no longer final. Hence -e is not lengthened before the plural endings -ha and -gdn either. Similarly, the proclitic preposition j he “to" is never to be scanned long; neither is ze “from”. However, -e in the first member of a compound is capable of lengthening as may be seen in §84 first example (ddnekas). Il should he noticed that in the Classical Persian orthography there is correspondence between indicating and riot-indicating -c by the letter * on one hand and capability and incapabilii) of lengthening on the other. Thus, a,L>. sdn? “house \ hacce “child‘\ na “no", ^ it' "what \ *£* ke “who" etc. sometimes have to be scanned \ane, hacte. na\ * ke etc, whereas in sunefm “houses'', btucfgdn “children", j be “lo*\ j ze “from", \ x eera “why" etc, -e and -v are never lengthened. In the Modern Persian orthography this correspondence no longer exists, be "to" is written and xanehd has become §71 -e is sometimes contracted with ezafe to form a single long syllable which we transcribe ee, as in the following well known lives by Daqiqi: he do ci- ■w r ■ giran- yek I par- vek i zar- * 9 m degar 1 8 J- nidni 1 5 r ltl e ndm e hart e d- d’ mar mam vek i raV- m* * * malek bar fr’diide e lakat ra farani neheste 1 7 *■ vamani 15 Cf. §92, 16 Cf. §112. 17 For ihe short e cf. §41. 18 For Ogar. cf. § 118 . 32 THE SYLLABLE §71 "With two things they gain (a: one gains) a kingdom, one silk-blue, one saffron-yellow: one (is) gold with the king's name inscribed thereon, the other tempered Yemenite steel’. This contraction is quite common in the poems of Moulavi. Elsewhere ~ee is extremely rare. (It occurs in sahname, 11,244,986 and VI,86,31l; Bertels’ edition.) §72 Similar contractions are heard in colloquial Persian where for sane >v md "our house" cme may hear \3nema etc, En reciting poetry -e v may be similarly pronounced* bur traditionally it is pronounced eu and this is preferable* for -e? goes back to Classical Persian Hiji which must have been pronounced as a diphthong at, As mentioned in §3 at yields Modern Persian ct Hence the last mesm V in the above example should be pronounced: ik’ftitr a - hun e a- yamani li * §73 The metre of the last example is slightly different from the metre of 4 sahname since the last rokn in each mesrd 9 is not apocopated. It is fa felon v-- like the other arkdn, not fa^al (or fa^ul) v as in sahname. Hence the metre is said to be salem "sound”. The metre is then hahr e motaqdreb e mosamman e salem , generally referred to simply as hahr e motaqdreb e salem. Cf. §27 (and §§32 and 42). §74 Lengthening of final -o in do except of course before enclitics 20 . A . A- ** / \ zamdrte be xun e m * bar anda- m e to mii- ‘two” and y to "thou” is frequent, to tesne savad y desne savad m -w 19 Hubert Darke's article A Persian Prosod Uni Problem (Iran, Journal of the British Institute of Persian Studies XIX, 1981. pp. 151-157) quotes more than twenty examples of f e from Moulavi as well as a few from Naser Xosrou. Xaqani and others and gives a detailed and thorough discussion of the problem, but I cannot agree with the author's conclusion “that in this single instance Classical Persian prosody allowed two light (short) syllables to take the place of one heavy (long) one", which would mean that we should scan degar ahan e ab'ddde ye yamani ^ w •- against the rules given in §§150 and 152. (Hubert Darke also quotes some rare examples of rr and eo. Cf. Additional Notes ■ ■ Sab b v below.) 20 Before enclitics ) may be omitted lyor by lord “thee", c— i y jts". . .. .^or c ; kar e to si “it is your task". Cf. §§70 and 99. PINAL. SHORT VOWLLS 33 $76 "Fate becomes thirsty for your blood; to your body (even) a hair becomes a * ' r . In this be it both to and to occur. An example of do was seen in $71, and do occurs in the following heir. r** - r u +# do faskar be ham an- dar dvi- x'tand ■ '■ to gojti be ham an - dar ami- x 'tand “The two armies came to grips with one another; you would say, were mixed up completely". they §75 Lengthening of y? eo "as" is extremely rare, the poets preferring the synonym j cun. jfc. // #* *» ** * yek / gd- i” didam eo xorram bahdr sardpd- y* neiran- g o rang d negdr "I saw- a cow like the delightful spring; from head to foot a miracle and grace and beauty". This example shows the ordinary value of eo. The following be it contains one of the extremely rare occurrences of eo: jjC\ * I «4 #* namdnad be sar eo bad o ni- as- k ‘ bar hi ■ t o has c kas "From one end to the other the world is like a fairy tale and that's all. Good and bad have permanency for no one". §76 It is characteristic of nuzm v rurdn "fluent poetry” that lengthening of final short vowels is scarce. According to Elwell-Sutton (p, 213) -o occurs approximately one and a half limes as often as -o in t'erdousi and Mouluvi. -6 being slightly more frequent in the latter. In 21 Cf. §83. 34 THE SYLLABLE §76 Hafez -o occurs five times as often as ~o. -e :: is eight times as frequent as -c : in Moulavi and iweniy-one limes as frequent in Fcrdousi. In Hafez -e does not occur at all in the material examined by Elwcll-Sutton, but his material is too scanty 2 \ and does occur in Hafez, too, My own counis yielded the following results for Hafez; In 495 obyat a total number of 327 short -e. 10 long -t\ 305 short -t\ 0 long h\ 310 short 68 long - d . It is seen that -e is very rare in Hafez and that -e does not occur at all At I he end of a mesra V. however, -v and -e (to be scanned -e and -e) are quite common. This is a very strong argument in favour of my view, that the final syllable of a mesra is anteps, i.e, indifferent w r ith regard to length. (Cf §$41-43). Jt is difficult to believe that Hafez, being so reluctant to lengthen -e elsewhere, should readily lengthen it at the end of a mesra ^ not to speak of -e! On the basis of the facts stated in this paragraph and in §§57 and 71. we may arrange the modifications of final short vowels in order of increasing sang ini "heaviness* \ First comes P (ezate) as the least sought - it is followed by 6 and v>; then comes ~e followed by -e and -a:, and finally as the most xtrngm "heavy' 1 comes -ee, §77 Before ast “is" a final -e may disappear completely. In this ease the alef is generally not written and the he indicating -e is often omitted, too. Thus, touhe “repentance" + ast when pronounced touhast may be written in three ways; I J .y or ^ 1 eUafj 99 . / ’j) no horn bit- b e toubas- t o rah e sa dahom dar mondjd- t o x atm e ketdb The ninth is the chapter of repentance and the way of righteousness. The tenth chapter (contains) prayers and the conclusion of the book”. SaSdT In the following example move "amount" is contracted with ast to mdyast. M ** i* bozorga" 2 * koddman- d o dastu- r ’ kist ce mdyas- tesd" 2 * gati- j o ganjit - r ’ kist “Who are the nobles and who is the vizier? What is the amount of their treasure and who is the treasurer”? 22 Including -e and -a, which Elwell-Suiton does not distinguish from -e. 23 Cf. S $ I (with fooinoies). 24 Cf. 583. §79 FINAL SHORT VOWELS 35 i he disappearance of -e before ast is especially common in the perfect tense. Here the poetic language agrees with the colloquial language, where karde-e { j! ance, ince “that which ’, - * T - *. and harce “all that" etc. may when the metre requires become ^_Sv' 27 Cf. §§92 and 93. 28 Cf. §83. 29 To be analysed b-andazud-a (for be-y-anddzad-d): may also be analysed and pronounced he-ndazad-d. Both forms are heard in modern colloquial pronunciation. Cf. Lazard §23,7. 30 Cf. §§92 and 93. 31 Cf. §86b. 38 THE SYLLABLE §82 ank, k__Sot ink - one, & I ine, and hare etc. These forms, which are also found in prose, may even occur at the end of a mesrd ( i. W * • I • *• bar d"c 2S az gel dmad eo besna- .x’tand i2 sabok xe.s- f 'rd kd- i'bod sd- x 'tand “When they realised all that which came from (o: all that could be made of) mud. they quickly made a mould for mud-bricks ’. to az daryd jodai yo tojab dnk ze to vek lab- ze i n 2 8 daryd jodd nist “You are separate from the ocean, but (lit. and) strange (is) it (lit. that) that this ocean is not one moment separate from you". Sana ! The metre of the last example (., § 194. w — ) is described in 32 The form hesndxtand for bexenaxtand preserves an older pronunciation. Cf. Pahlavi snaxtan “to know” and §§102 and 106. IX NASALIZATION §83 The letter j when following a long vowel and preceding a consonant does not affect the metre. In other words -UJ~* JU- jdn midehad “He gives (his) soul" is scanned like L>- jd midehad “He gives place" — v not like *l>- jam •’ midehad “He gives the cup" - Whenever j does not affect the metre, we transcribe it " {jd" midehad). Thus **\ t ** *** ** *• Ay * ♦* • ** ' bidmu- x'esd" res- tan d id- / ’tan he tar an- daru " pd- d* rd hd- f’lan “He taught them to spin and twist, to weave the woof into the warp”. §84 If j after a long vowel and before a consonant appears to affect the metre we may be sure that it is to be followed by an ezdfe as the second jdn in the following example : maydzd- r > mur i ke dune - kas 2 ast m t J A 4 ft ke jd n dd- rad 6 jd- n e siri " x"as ast “Don't molest the ant who pulls a grain (lit. who is a grain-puller), for he has a soul (or life) and the sweet soul is precious”. In the following beit only the correct scansion can decide who is cursing whom (Zahhak is secretly watching Fereidun and Sahrnaz): t For be\dmu\! One cannot help asking whether the fact that n is left unwritten before consonants in the cuneiform inscriptions of Old Persian has a similar explanation*. §87 In the word j 2 in ha r (zinaharl) “beware, protection" the n is always reckoned in the metre: 6 Cf. R.G. Kent: Old Persian, §39, New Haven. Connecticut 1953. Dr. Jamal Rczai informs me that the xorasant dialects, too. show nasalized long vowels, thus no "bread” and jo "young”. 42 THE SYLLABLE $87 he gar nr- s ’ti an- Jar asta w o zand feres id- de rd zi- nhdr az gozand 1 "For had there not been according to Zend and Avesta immunity (lit. protection against harm) lor the envoy. If the metre requires no short syllable, the form zenhiir is used: 'Vif' fr. ■M"/ ! rd he jd” dd- negar ta bidr i 7 8 “He spared the life of one and said, see that you bring your head in concealment (o: that you get away and don’t show yourself) Before the suffix jj - var j sometimes has to be reckoned in the metre. Thus, a word like jdnvar , which is ordinarily scanned jd*var sometimes has to be scanned jdn'var — ^ — (or perhaps rather jdnavar , cf. the synonym jdnavar and § 1 18). he jdn i k az u ja- n 'var sod haydt he jd"dd- var i k a- rad az qam najdt "(I swear) by the soul by which life has become alive, (and I swear) by the Redeemer of souls who brings liberation from grief’. Nezami The word jo- “how. why; as"— always scanned as one long syllable- is read con as well as cu". Before a vowel it is cun. $88 After a diphthong and before a consonant j does affect the metre, but the case is a rare one. There is not a single instance in sahndme. An example is seen in §257 first heit. 7 The modern pronunciation of this word is gazcmii . 8 Mart for hr van, ct §93* X FINAL LONG VOWPXS §89 Wilh Ihc exception of d, all long vowels and diphthongs may be shortened before vowels including ezdfe. Shortening of u (and /7) is rare. ifj} • ** * •* * *» * * Vfi yek i jd- du umad be (find- van be ird " be do V vd ye peiqam - ban “A sorcerer has come with a (new) religion to Iran claiming prophethood”. Of a similar nature is the shortening of u (and d) before suflixes beginning with a vowel: honor x'd- r* sod jd - dui or- j'mand neha" ra~ s'ti d- s'kdrd gazand 1 "Skill (or virtue) became despised, witchcraft appreciated, truth (or justice) hidden, mischief (or injustice) open”. §90 Short i (or /) and long / (or /') appear to be equally common before vowels, whether these vowels belong to distinct words or to enclitics. In the previous paragraph there were two instances of long ~i before vowel {jadui and rds-'ti ) ; in the follow'ing example -i has been shortened. AS A j J s) If pas dgd - hi dmad be mddar ke farzan- ze farrox pesar d' sod ta- j yar I Classical pronunciation gozaml 44 THE SYLLABLE §90 “Then news came from the illustrious son to (his) mother, that (her) child had become king (lit. crown-wearing)". The following examples show long and short values before enclitics and in proclitics. zaman i Jehad tax - t o tdj b koldh zaman i qam b x'a- ri yd 1 ban - d o cdh "At one time it (o: the changing fortune) gives throne and crown and tiara, at another time sorrow and distress and bond and dungeon". bezdd b be saxti v bed a n zi- s 'ian zd- M vo 3 naka- m ' zist -■r r •' bdyad gerist "He was born and lived in hardship and disappointed. One ought to weep bitterly over that life". SjUL'S irj. (f !/y to ra bit - d ’ bdyad hand pi- s ’rou ke man raj- taniam to said- r e non “You must be the general, for I am (soon) to pass away, (and) you (will then be) the new ruler". (f )/ 1 v It zamdne biandu - h ■’ gast az badi gereftan- d* har kas rah e /'- zadi "The age became unvexed by evil. Every one took the divine road (of righteousness)". 2 Cf. §55. 3 Cf. §55. $91 FINAL LONG VOWELS 45 cfi bis JZsy * .. be jandal coni" gof- t J stih e yaman ke bid - farina t mabddd dahan Thus spake the king of Yemen to Jandal, (that) ‘may your mouth never be inauspicious’ The following example is taken from Nezami. ** pas az 'ti- farm a- farinan- de rd ke bind - i u da- d’ hinan- de rd E* vek i yd 5 •at # mm bed it bar ■■ » vek i rd m ’■ 4 nitiz 6 vekti vek ar m / i hame xal - q ’ rd kd - r'sa z “After praising the Creator, He who gave those who see (their) sight, (who is) One and everyone in need of Him, the Helper of all creation one by one §91 To the traditional grammarians / is a sort of diphiong consisting of e (Classical Persian /) + y (cf. §34). The malju r “pronounced” version of the firs! example in §90 is therefore T yjS* k tyj m (f d* pasaga hevamad zefarrax ' pesar hemadar kefdrzan dosadia jovar To them, this is in no way different from the example quoted in §44 (detandar - ) and requires no special rule. In both cases hamzv is said to be elided [agahey amad becoming agahey amad. Cf. §§44-46). But a long t before an enclitic is. seen from the viewpoint of the traditional grammarians, an irregularity since an enclitic cannot possibly begin with hamze (cf. §47), In this case ^ is 4 May also be scanned pas a: a- fare a- farmon- de rd 5 Cf. § 55. 6 Cf. §93. 7 Note that the j in the malfiiz version is written jj ■ 46 THE SYLLABLE said to be doubled lor the sake of the metre, and in the lexis this is often shown by a umiid (') m W * §92 Words with an internal i before a vowel like OL* midn “middle”, ol*- siah "black” may conveniently be considered a special case of the above rule, but it should be noticed that in such words the / is never long. Of this we have already seen several examples: §64, §71, §79 first example. §80 second and fourth example. §93 From the point of view of Classical Persian ihe vowel / in midn and siah is different from ihe i of t\g. him "fear” and situ "silver”, and identical with the e in jS ye/ "day" and behest "heaven” (Classical Persian git and hihist). mian and siah go back to Classical Persian mi van and sndie That they have not become *mevdn and *.vt yah in Modern Persian* as one might have expected (cf §3)* is due to a special sound law 10 the effect that Classical Persian -/r- becomes Modern Persian (not Hence also the verbal prefix he * takes the form hi- in words like r*** hioftam "that I fall” | Classical Persian htyuftam)* w hid (pronounce hid) "come”! (Classical Persian hiytih cfjw hum "that you bring ’ (Classical Persian hiydri). hiuhim "let us get” (Classical Persian hiydhim) etc., but owing to the pressure of analogy heyoftam, he yd , heyan and especially heydhtm or hiydhim are also heard. Cf Lazard §21, L NOTE; This sound law does not apply to initial Classical Persian /!■-. cJM "state” is not *ia/au but eydlat or colloquially aydht. §94 In Arabic loanwords ending in -iat or -ie, the i may be long: jaha" mot - tafeq bar foru trtd"- de a~ kon- ■i r * * “The world agrees upon His divinity, (and is) perplexed al the un- fathomablcness of His nature”. SaSdT \d a hi- fi e muhi- vatas 8 The reason for the long i is that these words in Arabic have <_£ -it- : aZaJI 7 idhivvat and •* • • ♦* # / zar d neH- mat aknu" ke baHd az to biru" m “Give out gold and wealth now. when outside your command”. bedeh k a- n e to st ze forma- n e to st you have it. for after you (die) it is SaSdi to st is also written or C— J. The latter is commoner. 12 Namely that stone can be changed inlo silver. 13 A pun on the word razi : You have to accept both the teachings given in the preceding ahyat and your own lot. 14 -am is an enclitic pronoun syntactically belonging to pis, 15 Cf. §110, 16 The pun on surhcingi and hamdhangi and the ambiguous use of navd make this hi’it untranslatable. 50 THE SYLLABLE §99 After final -e elision of a is less common. Examples have been given in §§77-78. final -e and -a are contracted with ast to -ist and ~fst respectively: ce + ast becomes d cist "what is?"; no -t- ast becomes C ; n ist - * . ♦ tab' ^ ' A is not . Such contractions fall outside the scope of this book and are treated in the standard grammars. See e.g. Lazard §§100, 101 and 157. Note that in some rare cases the metre requires C— w to be read navast. •r § 100 Very rarely the a o’ ast is omitted after a long vowel followed by O I have only met w'ith examples of this in the works of Moulavi 17 . {$1 * • z • •• ** / j £ S3 do tlaha" da- rim * guy a hamco net yek daha” pen- ha" s'-' dar lab- ha ye vet “Like the reed we have two mouths; one mouth is hidden in His lips" For the metre of this example sec §201. For the sense compare § 170 last beit. Persian docs not admit of triple consonant dusters, and even if the poetry ol' Moulavi is known to be sunjiin "heavy”, it is not probable that he should have produced verses with such clusters. Examples of the above type are therefore very sirong evidence for o being realized as a supra segmental nasalization in the position between long vowel and consonant, since with that pronunciation the queslion of triple consonant clusters no longer arises. The Classical Persian pronuncation of C — must therefore have been pinhu st. Cf §86. 17 See e.g. koUtdt e sums, lines 20866 and 32148 (.w/wit t* tli" .tv). XI ALTERING THE SYLLABIC STRUCTURE OE WORDS BY SUPPRESSING SHORT VOWELS. DOUBLING CONSONANTS ETC. §101 In ordinary prose the enclitic pronouns have the forms - am “my", -at "thy", -as “his etc,", -emdn “our", -etdn "your", -esdn “their" when following consonants, and the forms -yam, -vat etc. when following vowels. In the spoken language the forms -m, - 1 , - s , -man, -tan, -sdn ] generally take the place of -yam, -vat etc. after vowels (Cf. Lazard §91). In the language of poetry the forms -m, -t etc. are freely used even — with the exception of -m — after consonants. Hence u I ( JLob. “I gave him water" may be scanned in four ways: dddamas db - w-, dddamas db - v - dddams db , dadams-' ’ db — ^ We have already seen several examples of the ordinary forms: §36 second and third examples; §60 first example; §65 third example; §77 second example; §79 last example; §83; §84 second example; §90 sixth example; §94; §97 both examples; §98 second and third examples. Compare those with the following examples: .A _ * * •• • / » v co bd mar- d e ddnd - t ' hdsad nesast zebardas- t' gar dad sat e zi- r'dast “When you happen to sit together with a wise man, (your) inferior head becomes superior (i.e. you learn a lot)". be vazdd" * ** be dels an- har d" kas dar dyad ke sod na ze bar su sepas hards l Or rather -mim, -tun, -sun, cf. §3 footnote. 52 THE SYLLABLE $101 “Whoever has become ungrateful to God, from every quarter dread enters into his heart". t co ra/'ti sar a kd- r * ha zad ast agar nt- k ,f basad- l' 2 3 ka r ar bad ast “When you have gone, (your) business is with God, whether your work is good or it is bad”. konuHd” bebavad m bar e u * sodan be bar bh * * s o kam rd~ y e farrox zadan be xubi * soxanhd- s* pdsox dehid 1 co porsad soxan rd- y e farrox dehid' m "Now you must go to him and give (your) fine opinion on everything, great and trilling. Answer his words well. When he asks about a matter, give a Dne opinion". Other examples may be seen in $230 first beit. §231 second beit . and §241 second beit. The following beit contains two of the extremely rare occurrances o -m "my" after a consonant. * * * * ♦ bn nax'dnam nabarde barddar - rre' ra nasuzam del e pi - r* rnddar- nr * ra "1 shall not call my warlike brother; I shall not torment (lit. burn) my old mother’s heart”. •v w §102 A sequence KVKVKV- 4 may be changed to KVKKV-. In other words 2 In present day recitation pronounced basal!, in the lime of Ferdousi probably hd\abr- 7 K 3 Bertels’ edition has nehid. 4 K = consonant, V = short vowel. V = long or short vowel. § 102 ALTERING THE SYLLABIC STRUCTURE 53 two initial short syllables may be changed into one long by eliding the second short vowel, thus safaqat “compassion, kindness" may become sqfqat as in the following beit by Nezaml. setam ra be safqat hadal kar - de niz has i mos- kel e dah - r* ha l kar- de niz # m “(I have) also changed tyranny to kindness; (I have) solved many problems of the age as well". This elision is especially common in short syllables following the verbal prefixes be-, ma- and na-. Thus bederavi “you reap" may become bedravi and besenavi “you hear" may become hesnavr, manesan “do not place!!" may become martsan and nanegarad “he does not look" may become nangarad : negar td ce kdri hama" bed- ravi sqxan har ce giii hama n bes- navi “Be careful what thou sowest (for) thou shall reap the same; whatever word thou speakest thou shall hear the same". makan x H a- r' x'ahan- de darvi- s' ra * bar e tax - v ' mansa" badandi - s ' rd •9 “Do not despise the suppliant pauper; do not place a malevolent (person) near the throne". be giti har a" kas ke ddrad x 'arad su-ye mar- dom e bi- navd nan - garad “In the world everyone who has, eats, (and) does not look at destitute people". 54 THL SYLl.ABLL $103 §103 If ihe vowel elided is o, be- is changed to ho-, whereas ma- and na- remain unchanged. Thus begosdy “open!" becomes bogSclw but magosdy “do not open!" and A»LL5 sJ nagosdyad "he does not open" become nuigsay and nag sov ad respectively, and ydZj besenou “listen!" becomes besnou (not *bosnou) since the vowel elided is e not o. feres la- de ye sal- nr' cu" gas- t' bdz sahansd- h ' bensas - l o hogsd- d 1 ' raz “When the envoy of Salm returned, the emperor sat down and disclosed the secret". There are few exceptions to the above rule. Thus, although begosud “he opened" becomes bogsud, still lyy herohud "he snatched" becomes berhud (not *borbud). fit/-. .. * Sjhf/ ** y,)t # qalam dor kaf e ti- r' beskas- tami fk koldh az sar e ntd- h * bet bit- dami * “I would break the pen in Mercury's hand, I would snatch the hat from Luna's head". § 104 Such elisions are freely used in verbs. I'hey are not uncommon in Arabic loanwords, but they scarcely occur in proper Persian nouns and adjectives. jJU be pedar "to father" does not become *be pdar , neither does fjJj pedaram “my father" become *pedram. § 105 Similar elisions are very common in the spoken language, where one hears namidt for nasvnhti “did not you hear?" etc. Cf. Lazard §12, §106 These elisions may have been caused by the "principle of least effort", a factor which is always at work in languages. However, since their occurrence is determined not only by the phonetic environment, but aJso by lexical considerations, it is reasonable to suppose, that the forms with the second vowel elided represent an older linguistic stage. In Pahlavi initial consonant clusters are common: bridati “to cut", drudan “to reap". framudan “to order", f ref tan "to deceive", frestadan "to send", fro. \ tan "to sell", griftan “to take", srudan "to sing", stadan "to take", ikastan "to break", sndxtan “to know", xridan "to buy", etc. Classical and Modern Persian docs not allow initial consonant clusters, so the ALTERING THE SYLLABIC STRUCTURE 55 S 108 above words have become boriddiLJijj* dor u dart (and Jjv yj* deravidan). jiv*/ fanmukm* fenfutn* ttTcsiadinh fimlxiun. \*ercf tan. sarudan* Jjlu~ setadan* xekastan , J xcnaxintti J-c vflrwbtf. etc\. all with the same meaning as in Pah la vi. In internal positions, however. Persian does allow consonant clusters. Therefore, though Pahiuvl xrad "wisdom" has become Persian xenui the form xrad is preserved in hexrad "wise". Similarly, forms like he (rest "send!", nafruxt "he did not sell", Acyrr// "he look", uiuvAtftf "do not break!" etc. may be original, in which ease beferest, uuforuxL hegerefh nuiscktuh etc, should he considered secondary forms made by analogy with the prefixless forms. Forms like herhitd and hogsaw loo, would have to be considered analogous formations, since the corresponding Pahlavi verbs do not have initial consonant clusters: mhudun and n isddan. §107 There are certain indications that initial consonant clusters did still exist in Persian as late as the twelfth century A.D Otherwise it is difficult to understand why Sums v Qcis , the author of ul-moSjam \ should have been so eager to prove that they do not exist. His argument is simply that human beings are not capable of pronouncing initial consonant clusters; C J ^ *Ija4 ehtedd he hart <' taken ..... maqdur e hasar nisi. He quotes the following words JU# faqdn "alas", fp deram "drachma". saray "house", somdr "count !". said by a certain JL_i Vj yl Lhn e Dorostuye ye Fasai from the province of Pars to be pronounced fqan % dram , sray and smfir respectively. The second and third of these go back to Pahlavl drahm and srfn\ Asa matter of fact one does occasionally meet with initial consonant dusters in the earlier Classical Persian poetry . An example — Jnisad "he sells" may be seen in the sixth heii of the poem quoted in §147, § 108 A sequence KVKVKKV- may be changed to KVKKVK.V-. Thus gorosne “hungry" may become gorsene: kabohstan “country of' Kabol" and go/estdn “garden'' may become kdbo/setdn and golsetan etc. J>£jfcJLS v •/ kas i k il naddrad bovad xos- k dab * 9 cona" cu " tot gor - sene ni- nr' sab "He who does not have, becomes dry-lipped (o: hungry), just as you are hungry {now at) midnight". ## 5 The earliest treatise on Persian prosody still extant, written 1217 A.D. See page 36 in the critical edition Qazvmi and Rczavi, Tehran s.a. 56 THE SYLLABLE § 108 par as tun- dega" rd su-ye go /- set an feres tad hand md- h e kdhoi - setdn “The ‘moon' of the country of Kabol sends the servants to the garden ". •• / l I Ji f lj& co esjan - didr d" xoddvan- d e ouran- “When Esfandiar, the brave hero, the ruler (lit. lord of the throne) of terrible appearance (heard ...)”. §109 garvsne goes back io *gorsne as shown by the colloquial form gain? (already in Pahiuvi ; gusnag. cf, the synonym gursag , both meaning "hungry"), When the language could no longer tolerate the cluster -rxn-, anaptyctic vowels made their appearance and produced the contending forms gomsne and gorsvng. The former was victorious, gorxvne being preserved only in the language of poetry. The same explanation mutatis mutandis can be given for got cstanlgo !se uin etc. Curiously enough the original forms *gotstan, *gtirsnp etc. would fit as well in the metres — gars'ne and got s' tun — but are rarely heard. §110 In Arabic words doubled consonants are very common, but such doubled consonants do not come naturally to the Persians and in ordinary speech they are very often pronounced single. Cf, Lazard § 19. The poets take advantage of this to let such words have single or double consonants as the metre requires: hami dor- r ’ ban- d' bar xd- k e xosk ' i hami a- mad az bit- s 'ta " bu ye mask “Pearls (o: raindrops) rained on the dry ground. Smell of musk (a: sw eet smell) came from the garden”. Here dorr is pronouned with double rr as in Arabic. *• ** r ** . ♦ • w •« V gov** e tah- g e hd sah- motan m o tan 6 Pronounce gow e, cf. §95 footnote and § 1 18, Some prefer to pronounce guv e. 7 Or tahm'tan. The ordinary form tahamtan was seen in §31, where as an epithet of Rostam it functions almost as a proper name. It is a compound of * tahm “brave" (cf. Old Persian laxma- “brave") and tan “body". 8 A hendiadys: “with dread and body" i.e "having a terrible appearance". §111 ALTERING THE SYLLABIC STRUCTURE 57 noxost a- lat e jan - g ' rd das- / ’ hard dor e net - nr' jostan be gorddn sepord "First he put his hand to the implements of' war. He entrusted the pear! of seeking fame (lit. name) to the warriors”. Here dorr has been simplified to dor. ?n the following be it by Sa c dl haqq occurs twice, once as haqq and once as haq. be haqqas ke td haq jamdlam nomud degar har ce didam xaydlam nomud "(I swear) by His truth, that since the True (a: God) has shown me (His) beauty, whatever else 1 have seen has appeared a day-dream to me". §111 On the analogy of alternative forms of Arabic loanwords like haqqihaq , genuine Persian words, too, have acquired alternative forms in the poetic language. The rare Persian words with doubled consonants occur with single consonants, ommid “hope”, farrox "fortunate” and JJo badtar "worse” may become omtd 9 , farox and batar ,0 . the latter via the assimilated form Jo hattar. hame naz- de man sar be sar kd- far and ' : v az dher- man 1 2 e bad- kones had • tarand "In my eyes they are all complete atheists and (are) worse than Ahreman the evildoer (who at least does not deny God s existence)”. 9 Perhaps amid should be considered the original form rather than ommid. 10 In this form the word has passed into Turkish: herer, 11 The correct pronunciation of" this word is kdfer, but the current form in the colloquial language is kdfur. kdfvr would here make a poor albeit permissible rhyme, so I have preferred to read kdfar , imagining that this pronunciation might date back to Terdousi’s time, 12 Cf. §1 19. 58 THE SY LLABLE $111 be gift beh az ra- s 'ti pi- se nisi ze kazzi 1 3 batar hi - c ' andi - se nisi ■ Hi ft ft “There is no craft in the world belter than honesty (lit. straightness). Worse than crookedness there is no thinking”. * j ** ** 9 f * ke cu n pu- r’ bd sah- m o mehiar savad az u ba- h •’ ra rii- z ' battar savad 9 m ‘{The Parsee said) that when a son becomes (more) formidable and greater, (his) father’s lot (lit. day) becomes worse on account of him”. In the above three ah yd l badtar , batar , and battar may be compared In the following beit bacce has been changed to have. pis d- yek i gii- baee pi- s az u ra/- te u ma"- de bud * f * * wt # * » “Before him came an onager. It was a female. (Her) colt had gone before her, (but) she remained”. To such forms the rule given in § 102 may be applied. Thus baecegdn “children” may become first bacegdn and then baegan : * m * * jay ' jdv ab- r e sepid an- hamco baegd- n e havdsel bar sar e dar- vd ravdn — — — ~ ~ “See here and there in the air the white clouds, (so) small, (so) small, just » like pelican squabs moving on the surface of water”! tJjj' Azraqi ] 3 For kazi cf, § 1 12, ALTERING THE SYLLABIC STRUCTURE 59 $112 The metre of this example will be explained in $199. Compare also the third example in the next paragraph (omhf). The form ommkl occurs in the quotation on the titlepage. §112 Similarly, single consonants may be doubled when the metre requires. Thus -r- very frequently becomes -rr- l4 . Stop consonants are rarely doubled in this w'ay. The effect of such doubled consonants ts sometimes astonishing as in the third example below, where jams id is changed to jamnv'sht. The first and second examples show doubling of final parr for par , and of internal darrandegan for darandegdn. The last example shows one of the very rare occurrences of a doubled stop consonant, eapp for cap. See also the examples in §§71, 111 second be ip 140 fourth heir, 147 last heii. 233 second sepdh e sab e //- yek I far- s ' ’ afkan - ’ Dark Night’s army on plain and the raven s plumage ". re bar das - / o ruq (U‘ cu “ 1 5 par- r c zdq slope had spread a cover (black) like y j 4* M 'J ze horrd- ye darran- degd" can- g e div sode SOS' ! ' az xas- m e keihd"- xediv ■ * "Because of the roaring of the wild beasts, the hand of the demon had become impotent in dealing with the fury of the world-ruler’ he gjii maddri- d'' eandi " omid negar id ee bad kar- d' hd jam - m 'std “’Don’t expect too much from the world, look at how ill it treated Jamsid ”! * 9 # * 14 The doubled - rr - is often etymological. Cl. Pahlavi parr ’‘leather*. and tkm 'ukm '‘to tear’ 15 Bertels* edition has gostank* az (and afgutuk* in the apparatus). 60 THE SYLLABLE $112 sab tabiha" jam V’ kard a~ capp o rant goft'jan e bar do dar das - t e soma st — w — ~ v/ “ * “The king assembled leeches from left and right. He said the life of both is in your hands". Moulavi This metre is explained in §201. §113 Such doubled consonants do occur in the spoken language chough rarely, cl'. Luzard 5 19 end. In poetry, too* thev are rather scarce and a form like fan un '\i- x"abanid" he laid low" 1 2 , we find x"dbanid VI. 103.551 3 . For JUljJsil angtzanad "he spurs”, we find angtzamd* Vl,104 T 562 + 1 We have seen instances in §60 second example, §64, §80 first and fourth examples, § M2 fourth example. 2 The translation given is here as elsewhere the one which suits the occurrence referred to. 3 I.e. sahname (Bertels’ edition), vol Vh page 103, verse 551. $1 18 POETIC LICENCE 63 For begorizdnad "he puts to flight", we find *jt hogrtzanad 4 , VI ,104,562. In the three examples above the causative suffix -an has been shortened to -an 5 . For naytiramid "she did not pause", we find _u*jU n-dramtd 6 , 1,17*3,570 var. 8. For nasomdh "you do not count", we find nasmari 1 , VIII. 110,967. For parddxte "freed", we find pardaxte , 1,31,48. In Pahlavi pardaxtan is intransitive and pardaxtan transitive/causative. In the bed referred to Jb _j beparddz "make free" also occurs. Hence what at first sight appears an instance of poetic licence turns out to be in perfect agreement with Middle Persian usage formally as well as syntactically 8 : az a" bad- kones dt- v ’ ru re zatnin P V I* - ' bepardd - z o pardax- te kon del ze kin “Free the ace of the earth from that maleficent demon and make your heart rid of hatred". For JU nasekarid "would not hunt", we naskarid 1,173.571. naskarid is the older form, cf. Pahlavi skarag “hunting". For and al^JbLi sdhansdh (i) "emperor( ship)" we find sahansdh etc., 1,101,367; 1,97,302. It is unnecessary to sdhansahi oLb give further examples of ah for ah. For gdvdn "bulls (a: warriors)", we find jf^S" gavdn VI, 77,189, and for jlS” gav we find & guv> w r hich in turn becomes gou (cf. §96), VI,] 10.651. Cf. the paradigmatic change between gav- and gav- in Avestan 10 . In Pahlavi the word was written with an ideogram which 4 Cf. §§102 and 103. 5 Cf. C. Salemann : Miuetpivsisch §95 {Grumbiss dcr Iranisi/wn Philotagie 1,1. Sirassburg 1895-1901). 6 cr. §80. 7 Cf §§102 and 10.3, 8 Cf C. Salemann: Minetpersisch §94. 9 Cf. §§102 and 103. 10 See A V Williams Jackson: An A vesta Grammar. §278. Stuttgart 1892. 64 THF syllable: §118 leaves the possibility open that a plural gavdn originally contrasted with the singular gav. For budam “I was”, ij; bud “he was" etc., we find bodam Jj bod etc., 1,171,547; 1,101,364. For hits ‘’mind” and jLiys husvar, bus id r “vigilant”, we find bos, jL-Lft hosyar, VI. 85, 285; VI 1,2 14,9. For kbit "mountain”, we find £ kob, VI,97,469. For xdtrtuS “silent" we find xdmos and xamos 1 1 , VI ,105,576; V, 1 4, 136 var. 24, and for y jdrdmits “forgotten" we find j foramos , 1,145,141. Cf. Pahlavi frdmos and frdmust. For bihude "vain” we find eJyj bihode, VIII, 1 10,967. For niku “good” we find neku, 1,35,31. For digar “other” we find degar 1,101.374. For ^ 1 tstdd i2 “stood up", -U-ob beistad 12 “he stands”, etc., we find jLuJ estad, beestad, etc., VI, 103,551 ; VI, 9 1,381. Cf. Pahlavi estddan # and estddan. For JJ^r- cinad "he picks”, we find Jlx>- cenad , VI,93,4Q3, cenad could be the older form, cf. Avcstan vicinao( , but Pahlavi has already cin~. For mirad "dies" and Ojyo birun “out”, we find merad and jj , j be run, more often pronounced morad and boritn , VI. 68,41; 1,113,556. The change of e to o is caused by the neighbouring labials (aided by the o of the infinitive stem and by the u in the following syllable respectively) 1 3 . For goubar 1 * jewel", we find gohar , 1,111,521, For I j s£3 ve * y d (Classical Persian vai rd) "him", we find 'jj va rd , 1,49,173. Cf. also §87. §119 Much less common is the lengthening of short vowels: For pdsox "answer”, we find pdsiix, VI, 104,564. For yd ahreman “the Evil Spirit", we find jg yd ahriman and /y yd dberman (the long a necessitates the transposition of r and e in the latter), 111,245,3707; 111,211,3218. I ) And xamus is tound in divan e hafez 10.7 (ed. Qazvini). 12 Or fstad btfsiad etc. 13 Similarly savtir "horseman 1 ' is sometimes pronounced sovar, and the prefix ne- be changed to no-. nemiuhn, nomudan "to show", rahnomun "guide ", etc, 14 Classical Persian guhar [go:har]. may §121 POETIC' LICENCE 65 For oobil oftadand "they fell", we find JU.sb&jl ilftadand , VI 11,220,2873. Cf. Pahlavi oftddan. For JL*I om(m)id “hope", we find y timid, IX, 334.289. For ostdd “master", we fmd tistad, VII .40 i . 1 708 . Cf. Pahlavi awes tad. For XU-il asoftand "they became agitated", we find -CJj ^ astiftand, VN 1,220,2873. For mehmdn “guest", we find mfhmdn, VI 1, 235,287. §120 Sometimes a vowel disappears altogether: For agar “if’, we find gar 1,99,336. For jlil anar "pomegranate", we find jl ndr, V.7,21. Borrowed into Turkish in this form. For azad “upright", we find jlj zad , VI, 69, 67. For paziroft "he accepted”, we fmd OijJL pazrofi , VI,69,67. For jy\ amir “commander", we find jy mir, VII, 113,7. §121 Much more common is the addition of an extra vowel: For ,J hi "without”, we fmd abt , I, 95,262. Cf. Pahlavi abe. For L ba “with", we find ll aba , 1,22,141. Cf. Pahlavi abag. ■ * For j> bar “on”, we find j>} abar , 1,60,161. Cf. Pahlavi abar. For *uiv ntise "fortunate", we find antise, VIII, 335, 338, Cf. Pahlavi andsag. For \ y j zird "therefore”, we find lyjl azira, i .16,58. From *az in ra (or from Pahlavi az ed ra, cf. §86b). For jlSw' dskar "evident”, we find FlSwl askarci, VI, 72, 101. Cf. Pahlavi dskdrag. It will be seen that in all the above examples the forms with extra vowels are really the original ones. For jjj^l avordan "to bring”, we find avaridan, 1,117,635; and for o*jj ji parvarde "brought up”, we find a-Aj jjj# parvaride, VI ,99,501. In these two examples we do not find archaic forms substituted for current ones. On the contrary the poet is here ahead of the development and employs analogical formations not yet accepted as standard. For OjJu j feridtin “Feridun”, we find jj-byl aferidun , 1,117,628. An instance of popular etymology. For Crd/d ahanin "made of iron”, wc find d/iartine , VI. 80,228. For pahlamn “champion", we find pahlavanu , 11,64,27. 66 THt SYLLABLE: §121 For l" . ajf goji "said", we find \’i>< goftd . VI,7 J .96. For jL bad "be", we find bu hadd, Vl,88,332. ■' * For j^>- .v an jar "dagger", we find I xanjard , VI, 98 ,48 6. For j~k>' andar "in", we find FjJl andara , Vl,98,486. For jjJ'JL segdiidan “to think of’, we find IjJISC* segdlidand, VI, 103,558. In the Iasi several instances a long d has been added to the word. Originally, this d was probably the asseverative enclitic particle d or hd, which is so common in the colloquial language- surprisingly enough not mentioned by Lazard but in the language of the poets it has degenerated into a mere verse-filler. As such it has been admirably parodied by JL^- JlTlj SObeid e Zakani in his j £ y mus o gorbe “Mouse and Cat", of which a belt is quoted in 5231. §122 Word initial VKK- is often changed to K.VK- and vice versa: For j-l5w 1 eskandar “Alexander", we find jXSC-. sekandar , VII ,6, 10 (Cf. VII ,6, 1 ). From Pahlavi *Skandar. For jlJL^- sepahbodan 1 5 “the generals", we find jlJL^.1 espahbpddn 1 5 , VI, 77,1 89. Cf. Pahlavi spdhbed. For abruhdyas “his brows", we find J-Ujy baru has 1 6 , IX, 327,209. Cf. Pahlavi brug. For separ "shield", we find espar, V, 199, 1 954. Cf. Pahlavi spar. For *jl5w sekdram “1 hunt”, we find » J>Jz\ eskaram, IV,313,176. From Pahlavi skar -, cf. § 1 18. The Pahlavi forms show all the above examples to be parallel developments of original word initial consonant clusters. In the following examples the poetical forms are the older ones, For JJz solar "camel'’, we find osior, 11,133,114. Cf, Pahlavi ustar. For j\ j fardz “high”, we find ji ^*1 afraz, 11,44,13 var. 2. Cf. Pahlavi abraz. For $ jZJd befesord "he pressed”, we find j jjJu biafsord 17 , 111,175,2685. Cf. Pahlavi ajsurdan . For seldb "haste", we find *_jUJil estab , 11,192,296. Cf. Pahlavi ■ aw is lab. 15 For the development had to had . cf. § 1 IH end and footnote. (The rhymes suppose had.) 16 Cf. § 101. 17 Cf. §93. POF.TK LICENCE 67 §123 For iS^j^ sepordi “you entrusted", we find espordi , VI , 1 1 1 .675. Cf. Pahlavi ahespurdan. In a number of cases the poetical forms are ahead of the standard forms in development. It should be noted that the vowel quality often fluctuates: For 1 afrttxi “he kindled", we find j foruxt or for fix 1,236, 1478, For Jjjil afzun “more", we find jji fbzun, VII 1,7,22. For qfzud “increased", we find j yp foziuK 1,101,376. For afsdne “tale”, wc find fasdtre, V, 304,1 1 56. For r *j^\ afstjs “sorrow”, we find # Jbstls , IV, 118,46. For afsun “fraud”, we find fosun , IX, 229, 3659. For JULijl qfsand “he scattered”, we find JjLii fesand , VI 1, 163. 1 45. For j&i of fad “it fell”, we find fetdd. 1,1 17,637. For Xxil afkand (Classical Persian afgand) “threw”, we find xSi fegand' 8 , 1.106,453. Classical Persian afgandan (Pahlavi abgandan) has become afkamian in Modern Persian with assimilation of fg to fk. whereas in fegandan g remains. §123 Sometimes a consonant disappears, If the consonant is followed by a short vowel (motaharrek), the vowel disappears as well: For j\j\ dvdz “voice”, we find 1^1 dvd, 1,171,544. For gg Jjj i farvanfin “March/April”, we find farvadin , 1,42,53. For § | agar “if', we find jl ar , VI, 70,82. Cf. §60. For jjjl avarad “he brings” and jj \ avar “bring” etc., we find ijl drad and j\ dr etc. 1,106,459; V,7,I8, This change is very common in the colloquial language as well, cf. Lazard §23,2. Not surprisingly, h is the consonant which most often disappears: For a\S govah “testimony”, we find 1^5* govd, VI, 69,68. For oLfT gidh “herb”, we find \S gid , 1,15,52. For gy* hie “(none) at all", we find ^ fc, 1,16,58. For Jp. cehel “forty” and cahdr “lour”, we find Jr cel and jU- edr, IX,225,360I; IX, 308,8. Both are common in the spoken language, cf. Lazard §23,3. See also second example in §116 18 Bertels' text has probably a misprint. 68 THE SYLLABLE §124 §124 Various other changes: For zardost “Zoroaster ", we find C—iJtojj zardhost , VI, 70,82 (the form preferred by Daqiqi), and zardhest, VI, 26 1, 72 1 (the form preferred by Ferdousi). zardost , which is less common in sdhndme (e.g. VI, 71, 92), goes back to Pahlavi Zardust : zardhost! zardhest come from *zardohest from Pahlavi Zarduxst. For jl*j jfc har zamdn "all the time", we find hazmdn , 1,103,402. For busy dr, hwiar "wise", we find j \ ho.ii var, 1,13,24. For pesar "son", we find po.s, VI, 8 1, 241. Cf. Pahlavi pus and pusar. J&. Safv "pardon" must in 1.22.146 be read 'iafu, and qor an with the proper rhythm” is more profitable than the study of Arabic grammar. 70 THE SYLLABLE § 1 26 Enclitic pronouns ending in a short vowel ure lengthened when following a short vowel, thus "‘Hence Subutai set out for war towards Qazvin like a fierce leopard”. The reason tor this apparent irregularity must he that the name was really pronounced something like sohdtdy*. Since it is not spelled J >r - . sobuwi (Classical Persian pronounciation subotai), we must conclude that the sound changes described in §3 had already occurred by the year 1300 at least in the dialect of the author’. French and English loanwords are treated in a similar way by the modern traditionalists. 0 3 r e qorra " be jang poking 4 Mongolian sources give his name as suhegeiai > sube'ptai > suhetai pronounced suhetai (see Erich Haenisch: Worterbuch zu Mangho! tin Siuca Tobca'an , Gekrime Gesthichfe der Mongolcn, Leipzig 1939, p, 181), so the Persian pronunciation was probably at first Habitat, which leads us to the same conclusions about the sound changes as sobotui does, I am grateful to magistcr K£re Thomsen Hansen for drawing my attention to these facts, 5 On the other hand, the Demotic Greek phrases occurring in kollidt e tarns lead us to conclude that the same sound changes had not yet occurred in the year 1250- at least not in Moulavfs dialect. We thus read po.\ ise (jra*; doui) “How are you?” (line 24028). l s? JIT kali mera .ras(KaXr| pepa oa<;)"Good morning to you " (line 34261), 4 Ue/a rJo (eJ-u (kb) “Come here”! (line 34262), ayapo se (ayan& ae) “I love you" (line 2698 1 (T ). XIV RHYME § 129 _u-l> j Cjj ^ ^ — - ' nazm ... soxan i st ke data ye vazn o qdjie bcisad “Poetry is speech possessing rhythm and rhyme" \ Therefore it is not out of place to glance at the rhymes employed in Persian poetry before proceeding to an examination of the various rhythms. § 130 In Persian poetry, rhyme means that two or more sections — each section consisting ordinarily of a mesrd or a beit — from a certain point onwards are identical in sound, but not in meaning. The identical portion must contain at least one vowel and may have any length from a single letter to several words. (Up to §139 the examples are all with the exception of the one in §131 taken from Nezaml.) co vaqt e z a mane ft condn af- ke bar mur- rain! d- bar a rad sarad rii- ft g e x“is d- vad az ran- ft hahane * z 'garas i a das d- I o dard be mard * geld r'zu “When the time for departure (to the other world) comes, time brings to man a pretext of trouble and pain; lime squeezes his throat in such a way, that he desires his own death". In the first beit the identical section is -ard: in the second beit it is shorter still, being the single vowel -u. Still the two -us are not identical in meaning since they cooperate with other sounds in conveying the different meanings “throat" and “desire". 1 Jiilal HomSi: jjj ft man c hataqut o scndHat c adahi (Teheran 1354)* p- 5 74 THE SYLLABLE §130 In the next example the identical element is much longer: -[vide am • 7 be bar id sar az dd- ke raft an d o ddnes i- de am “Wherever I have been prepared to go, I have not turned away my head from justice and wisdom". §131 In classical poetry the majhul vowels / and i) originally pronounced e and <>- do not rhyme with the mttfruf vowels i and u. It is not a coincidence that the first long vowel in both puslctdf and nap h ide is majhul and the second vowel in both words ma^ruf. Cf. §2 (Post -classical poets, of course, rhyme / with / and u with m.) ■+■ * The most conspicuous efTecl of this is that the indefinite article / rhymes neither with the suffix -/ forming adjectives and abstract nouns nor with the enclitic verb -/ "thou art". Note that the Modern Persian stress-accent plays no role^ 4^ M// by 9 4 bell u 20 f- ! f tur ur to az out ivra bar nehadi koldh e "To him Tur said, ’ll’ you are smaller than we. why have you (o: crown) on your head?’ " kehi nwhi the hat of greatness Ferdousi Here kehi rhymes with nwhi. §132 In the traditional learning, rhyme is the subject of a special discipline *jii ^ fetm <■ i qafie "the science of rhyme\ possessing an elaborate and to my mind absolutely superfluous terminology. I has in the last example of §130, this "science of rhyme" has a special technical term for each and every atom of the rhyming element: ~hhk> anu analysed as -e-v-c-e Y-ibe- or *wh' -il-a- -a-m, is described by no less than ten different technical terms, which the reader may find in Elwell-Sutton, p. 225*234. §133 In the following example the identical element is -dust. be azar - m e soft a- n e darvi- s'dust t m 4 be darvi- s e qaneS ke soltd n x"od u st mm § mm 2 Not basiade ! Cf. W. B. Henning: Sagdian Loan-words in New Persian. BSOAS X, p. 104. 3 As is often the case in great Persian poetry, this bed evokes — by the different combinations of the various meanings of its words— simultaneously a number of different images. It is RHYME 75 §134 "(I swear) by the shame (or modesty or sorrow 3 ) of the sultan who loves the poor, (and I swear) by the contented poor one. who is himself' a sultan”. Here the identical element, -dust, is not a rhyme to the English speaker, who defines rhyme as identity of sounds between words or lines extending back from the end to the last fully accented vowel and not farther. But according to the definition given in §130 it is. For the European it takes a while to accustom himself to rhymes of this type, but in the end he can become quite fond of them. They are often brilliant word plays as in the following example. ze duzax masou tes- ■ soxan dar behest as- “Don’t seek from hell a remedy for the those Four Streams 4 ”. t o a" cd - r' juy thirsty! The matter is in Heaven and The rhyming element is -juy, identical in sound, but not in meaning. If ■* was originally pronounced a, the identical portion must in the time of Nezaml have sounded edrajdy , meaning ’‘seeking remedy" as well as “four streams”. §134 The rhyme may comprise several words. In the following example the rhyme occupies more than half of each mesraS. ze ranjam dar dsd- yes drad magar bar i" xa- k* baxsd- yes drad magar therefore untranslatable, azarm has inter alia the meanings "shame", "sorrow", "modesty" and "dignity ', daryisdust may mean "loving the poor" as well as "loved by the poor". Hence azarm e sot tan v dar viidust may he taken as the sorrow of the sultan who fights poverty in vain; it may also mean his shame or modesty, because he knows that he ought to protect the poor belter, but it can also convey the meaning of the dignity of the sultan whom the poor love Moreover by the Sultan wc may understand God, in which case both meanings of dar vis Just fit equally well. One should not try to decide upon any one of these interpretations to the exclusion of the others. They all coexist in the verse, which may be understood on several planes at the same time. The four rivers of Paradise. 76 THE SYLLABLE $ 1 34 “Perhaps He will bring me from affliction to rest. Perhaps He will bring forgiveness to this clay". Here is the identical portion is -dyes cirad magar \ § 1J5 jU,^ tnubiuian "Guebres" is not supposed to rhyme with jy — . moskman "Muslims", because the identical portion, -an. in this case is identical not only in form, but also in sense, namely "plural ’. Similarly gait/an "to turn" is not supposed to rhyme with ntjirn “to go". Transgressions against this rule are not very rare*. §136 Before the rhyme proper there may be another identical portion separated from the rhyme only by a single letter. The reader may enjoy the effect even without knowing the technical term for it; cT^c/> j's bed i" rou - sani f ze birou- qani "Alas, that a lamp with this (o: such) brightness is going to go out for lack of oil"! Here the rhyming element is -ani strengthened by rou-. In the following example it is -dz strengthened by care-. arestu jahd"di- be bfed - regi md n - de ye i d- resdz d az ?' 1 cd re baz ■ "Aristoteles, the experienced provider of remedies, was lost without remedy (in his search) for this remedy (i.e, the remedy here needed: a remedy against death". § 137 The rhyme effect may be strengthened by the use of internal rhyme, as in the second mesrd ? of the first example in §130, where in addition to 5 In the traditional terminology only -dyes is termed qdfie “rhyme”. The remainder, drat/ magar, is called radif “refrain”, 6 A number of examples is given by Elwell-Suttofh p, 237, and in the fourth heii of §101 above Ferdousi allows sodan to rhyme with zadan. Cf. also §232 footnote 16. 7 Babayofs edition has an. Si 38 RHYME 77 the rhyme proper, there is an internal rhyme, the first rokn, zamane, rhyming with the third, bahane. The final syllable of an interna! rhyme practically always coincides with the final syllable of a rokn. This will be discussed more fully below, $$ 149 and 192. Cf. also $28 above. (Internal rhyme is called safi “cooing”.) § 138 The arrangement of the rhymes does not show the same variation as in Western poetry. There are two main varieties: Either there is but one rhyme repeated throughout the poem at the end of each beit (sec below $ 1 40), or else the first mesrd 9 of each beit rhymes with the second, and the rhyme changes for each beit , as in the following siah la siah di- dam i" kd- r'gdh -e rig e siah to be db e a siah 9 garam bd- z-' porsi ke cu" bu~ de am nema yam ke yek dam napeimu- de am # bed a " tej * l e vekrit - *r *■ ze mdnam at ke mord nadide jaha" rd hami jd " sepord jaha " jom- le didam ze bald wo zir hanuzam nasod di~ de az di V d ' sir a- na i" si yo ses gar hovad si hezdr * hanti" nok - le git yam saranjd- m e kdr “(Alexander says on his death bed:) From Black to Black I have seen this place of action (o: the world), from the Black Sand (the Karakum Desert * to the Black Water (the Black Sea). If you inquire of me, how I have been (o: lived), 1 must say that I have come no way (lit. I declare that I have not 78 THF SYLLABLE Si 38 traversed one moment). I am like that one-day-old infant who died (and) not having seen the world gave up his soul. 1 have seen the world, the whole, from above and below, and still my eye has not seen enough (lit. has not become satiated with seeing). Will I not — if these thirty six (years of my life) become thirty thousand — at the end of the affair say the same remark?" The arrangement of the rhyme is seen to be aa bb cc dd ee. This type of poem is called a masnavi. (Note the plural masnavidt.) It may extend to any length, and all long narrative and didactic poems are composed in this form. We might therefore translate it “narrative poem", hut the word, masnavi, itself is derived from the Arabic word for “two" ( OUJl iOnani) and indicates that each rhyme occurs only twice; we shall therefore leave it untranslated and employ the term masnavi. § 139 Another important characteristic of the masnavi is that with very few exceptions (see §§ 1 72 and 252-253 below) each mesra V always comprises eleven syllables. According to Elwell-Sutton (p. 245) “this strongly suggests a connection with the poetry of the Pahlavi books, where the eleven-syllable line is also the norm, though the metrical pattern is less clear". Also the Turkish kutadgu bilig “Auspicious Knowledge", finished in 1069/70, is written in eleven -syllable lines. See Chapter XXV. All the examples quoted so far have been taken from masnavidt with the sole exception of the example quoted in §71 and the last beit in § 1 1 1. where incidentally each mesra V comprised twelve and fifteen syllables respectively. § 140 The other main rhyme arrangement may be seen in the following poem by Rabe^e ye Qozdari. §141 RHYME 79 ze has go 1 ke dar bd- > ' -K q ma vi” gereft caman ran- g e artan- g e mani gereft magar ces- m e tnajnu" be ahr an- dar ast ke goi ran- g e roxsd- r e feili « gereft be mei md- * nai! andar kaqiqi" qadah sere.sk i 9 ke dar lu- le ma'vi* a gereft sar e nar- ges e td - ze az zar- r o sim ne.sdn e sar e ta- j e kasri 8 gereft co rohbd" xed andar lebds e kabitd banafse magar di- n e tarsi 8 gereft “For all the flowers that have taken refuge in the garden. the bed has taken the appearance of Mani’s Artang 9 . But is Majnun's 10 eye in the cloud, since the flower has taken the hue of Leila's 10 face? The tear (i.e. dew-drop) that has taken refuge in the tulip resembles wine in a jade cup. The fresh daffodil has taken on the character of the crest of Chosroes' crown of gold and silver. Like a Christian monk the violet has pul on a violet dress. Has it adopted the Christian faith"? The identical element is -i gereft, and the rhyme arrangement is seen to be aa xa xa xa xa 1 1 . § 141 This arrangement is common to the qazai and the qaside. (Note the plurals oLhp qazalidt and JL*Uai qasd'ed.) The qazai — we might translate it "sonnet" or "elegy" — is a short poem, rarely longer than fifteen abydt 1 2 , with a lyrical or emotional content as in the above example. "It is a marked feature of the form that the several abydt stand in no direct relationship to one another, so that they might be arranged in any order without affecting the general sense of the poem. All the same, although there may be no definable connection between the individual abydt , these ought never to be out of harmony with one another, and a single tone of mind should run through a whole poem" 13 . 8 -ti has been changed to -i for the sake of rhyme. See below §146. 9 The name of a wonderful book of paintings given to the prophet Mani by the angels. 10 Celebrated pair of Arabian lovers. 1 1 x should be understood as “no rhyme". 12 Qazai no. 1940 in ko/liat e sums comprises no fewer than 82 abvat, but many would not call it a qazai. 13 The formulation is that of CJ.W, Gibb. See his History of Ottoman Poetry vol. 1, p. 82 (London 1900). 80 THE SYLLABLE §141 The qaside . which we might translate “ode", is a longer poem. It is rarely less than twenty abydt and it may extend to as many abydt as the ingenuity of the poet allows, sometimes more than a hundred and fifty. The “standard" qaside begins with a lyrical or erotic portion, but ends up with praising the poet's patron and begging his favours. However, the qaside may have any subject matter, mystical, didactic, epic, etc. The introductory portion of a qaside is very much like a qazal — and therefore often called Jyu taqazzol — but without its discursiveness, that is, one cannot change the order of its abydt without affecting the sense. (The reader may experiment with the order of the abydt of the qazal quoted in § 140 and the taqazzol quoted in § 147.) The term qaside is derived from J^ai qasd “purpose", so E.J.W, Gibb and E.G. Browne translate it “purpose poem", and qazal literally means “love-making", but as both are defined by form rather than by content, they are better left untranslated, and we shall therefore use the Persian terms qazal and qaside. A poem with the same arrangement of rhyme as the qazal and the qaside . but consisting of only two abyai is called a dobeiti literally "two-!iner“ ; we might translate it “epigram”, but as with qazal and qaside we shall leave this as well as the rest of the terms in this chapter untranslated and employ the Persian forms. The robd^i, the so-called “quatrain” is a famous variety of the dobeiti ; it will be described in §§255-260. §142 As mentioned in § 139 the mesra 5 of the masnavi normally comprises eleven syllables. For the qazal and the qaside there is no such restriction. The mesra f may have anything from sixteen to as few as six syllables, though poems with less than ten syllables in the mesra 5 are rare. Very rarely the mesra 9 may extend to twenty syllables as in the example in §262. § 143 It should be noticed that in the first beit of a qazal or qaside both mesra ? rhyme. A poem where this is not the case, i.e. where the rhyme arrangement is xa xa xa xa , is called a AaU* qet^e “fragment”, since it is formally like a qazal or a qaside without the opening line. The qet*ie may have almost any subject, but should confine itself to that subject and not be discursive like the qazal. The reader is referred to SaSdi's golestdn, where he will find examples on almost every page. It must not be forgotten that the distinction between qet^e and qazal is in the last resort a formal one. A discursive poem with the rhyme §145 RHYME 81 arrangement xa xa xa.... is still called a qct^e, and a qazai may have a firm structure which does not admit of any change in the order of abyat 14 , § 144 Other rhyme arrangements, though not very common, do occur. For example, aa xa xa....xa bb cc xc xc....xc bb dd xd xd....xd bb ee xe..., etc.; aa aa aa....aa bb cc cc cc....cc bb dd dd dd....dd bb ee ee.... etc.; these two rhyme arrangements are called Ju > - j tarjtfband "return-tie” provided bb is a refrain which returns every time the rhyme changes, aa xa xa....xa bb cc xc xc....xc dd ee xe xe....xe ff gg xg.... etc.; this arrangement is called tarktbband “composite tie". aa...ab cc...cb dd...db ... etc.; aa...aa bb...ba cc...ca dd... etc.; these two arrangements are called mosammat “pearl-stringing” provided each section is of equal length, A detailed description of all these verse forms is outside the scope of this work. The reader may refer to E.G. Browne; A Literary History of Persia, vol. II, pp. 22-89. or better still to Jalal Homai's work on the subject, pp. 93-221. (See above §129 footnote ). Jan Rypka's short description of “the outward form (of Classical Persian poetry)" 'in his History of Iranian Literature . pp. 91-99, is packed with concentrated information, but it is nevertheless insufficient for the advanced student, and it is too difficult for the beginner 15 . As a first introduction to the subject nothing can compare with the introductory volume of E.J.W. Gibb: A History of Ottoman Poetry , which has been reprinted by Luzac in London and is still available 16 . It must be remembered that all the above verse forms and indeed all Classical Persian poems whether comprising two abydt or fifty thousand have the same rhythm from the first to the last be it. ♦ * # §145 The pronunciation of long a in Arabic is flat [se:] and quite different from the Persian a. In some Arabic dialects the long a has an even more fiat or rather closed articulation [e:], which made it liable to be identified with the Classical Persian majhul vowel t (pronounced e). A well-known instance of this is the Arabic conjunction valdkin or /akin “but”, 14 This is not so in Ottoman poetry, see Gibb, op.cit. p, 87. 15 Dordrecht 1968 (pp. 92-101 in the German version, tmnische Liferaiurgeschichie, Leipzig 1959), 16 The reader is warned that some of the technical terms are employed by the Turks (and by Gibb) in a way different from that of the Persians, notably the terms mosammat, *1L* hosn e maitaS and , *- hosn e maqtaS. w L. ^ 82 THF. SYLLABLE §145 which has been borrowed into Persian as jSLJj vafiken or liken , the former often abbreviated to eJsJ \ valik or vali. The Persian poets have exploited this quality of Arabic a to let it rhyme with a as well as with /. The change of a to / is called aJUl emale “inclination (towards the sound of i)'\ 's •• ^ * * na liar jd ke bini xat i deb .ferib tavani tama 9 kar- tianas dar ketih * “Not everywhere that you see a charming line 17 can you hope (to include) it in (your) book”. Sa^di § 146 A dif ferent variety of emale is to substitute -/ (not -f) for a final -a indicated by i£ with or without ate/ e maqsure 18 , as in the poem quoted in §140, where ma'va, leild , and kasrd have been changed to ma Y/\ leili, and kasri so as to rhyme with mam. Occasionally a final -d indicated by alef as in Lo donyd “world" may be changed to t£ -1 for rhyme’s sake. Thus in the same poem 1— J> tarsd “Christian” has become tarsi. The effect is to my ear not very pleasing. The opposite change, the change of -i to -a, is never met with. In the poem in §140 five final -as have been changed to T's for the sake of just one final The change of matti to man a would be unthinkable. EXERCISE : Before proceeding lo the second pan of this book the student is advised to read through and scan some hundred ah vat from a poem written in hahr e motaqareh. for example bustati e saScli. so as to become completely familiar with the mechanisms which bring about the rhythm. it will be noticed that knowledge of prosody often helps one to read and understand the text correctly. Thus the following lines from bus tan \1 xat also means "the down on the cheek of a youth", so that this heit means both that you cannot note down all beautiful lines of poetry in your notebook and also that you cannot form friendship with all beautiful youths whom you meet. 18 Of. § ! 27. §146 EXERCISt 83 might by some one unacquainted with prosody well be read hg del go/ lam a: mesr quad demand bar dustdn armaqdtu harand maid gar (ohi bovad az an quad dust w.xanhd ye similar a: qand hast. However, the prosody shows clearly that the correct reading is he del gof- tarn az me. i- r' qand a - vttraml bar e da* s'tan ar - tnaqattf harand mam gar tofu hu- d az a 1 ' qan* <1* dost soxanltd rt' she*- i* ■ tar az qan - d * ha.Kt ‘i said to myself (lit heart), from Egypt (people) bring sugar (and) take (it) as a gift for their friends. Though my hand was empty of that sugar, (at least) I have words sweeter than sugar". APPENDIX TO PART ONE: LIST OF PASSAGES QUOTED From Ferdousi’s sdhndme (crit. cd, Bertels a.o. Moscow 1966-1971): Page Verse number Quoted V in s Page Verse number Quoted in s Vol. I: 22 142 48 23 164 54 31 48 118 31 59 90 32 63 80 32 64 112 32 72 75 38 42-43 65 39 8 no 40 14 83 41 35 82 42 54 95 43 68 116 43 73 101 49 173 84 51 4 89 52 31 87 60 164 75 60 169 79 61 181 79 62 203 13 79 75 424 84 79 4 90 80 16 90 83 71 90 87 143 101 88 145 101 88 155 48 97 302 103 99 327 112 102 385 131 103 409 15 84 115 603 77 136 22 111 164 424 108 174 583 44 187 804 95 Vol. II: 64 34« 64 66 60 3 61 67 73 54 67 83 1 3 65 212 541 6 111 236 882 15 36 236 883 36 238 902 23 39 239 921 31 256 IV,8 8 36 Vol. Ill: 6 5 8 54 7 16 7 101 7 17 102 202 3078 54 Vol. V: 6 5 112 10 61 78 LIST OF PASSAGES QUO :T:D 85 $146 Page Verse number Quoted in $ Page Verse number Quoted in $ 20 206 80 24 270 54 28 342 1 77 86 4 65 88 36 80 Vol. VI' 9 : 71 86 97 72 106 89 78 194 116 79 214 60 79 215 65 80 222 87 94 420 101 104 573 80 no 651 !08 125 864 111 402 371 79 Vol. VII: 81 1367 127 190 612 102 316 202 102 316 208 108 373 1197 111 Vol. VIII: 146 1556 8 44 Vol. IX: 207 3328 79 Also from sahname , but quoted after Jules Mohl's edition (Paris 1838-1878): Vol. 11: 85 1154 74 284 97 60 318 903 90 Vol. VI: 141 1452 45 Also from sahname • , but quoted after J.A. Vullers: Firdusii liber regum (Leiden 1877-1884): Vol. I: 304 69 1 74 Vol. II: 823 918 90 990 509 55 From Nezami's eskandarname : Vol. I sarqfndme (crit. ed. ^Alizade, Baku 1947): 183 118-119 97 Voi. II eqhalndme 20 (crit. ed. Babayof, Baku 1947): 3 4var. 60 192 28 var.-29 129 19 The first ten quotations in Vol. VI are from the section of sahname believed to be the work of Daqlqi. 20 More correctly xcradnatn ?. 86 APPENDIX TO PART ONE § 146 Page Verse number Quoted in § Page Verse number Quoted K in s 193 56 102 194 72 130 195 76-80 138 196 97 var. 134 198 8 136 198 18 136 199 28 133 200 46-47 90 201 62 87 201 71 133 227 17 98 From SaSdi’s bustan (crit. ed . Rostam SAliyof, Teheran 1968): 3 16 94 5 I0 12 126 5 1 1 69 7 10 146 8 3 77 77 12 99 96 8 98 109 4 110 no- 9 127 183 10 98 190 3 98 209 12 19 145 The beit from Mostoufi's zctfurname in §128 is quoted after E.G, Browne: A Literary History of Persia, vol. Ill, page 96 (Cambridge 1 928). From Moulavi’s masnavi ve m masnavi (ed. Mirxanl. Teheran a h. 1374): 1 7 100 2 19 112 300 6 126 From Parvin E9tesami’s di van ‘Teheran 1974): 238 62 127 From Z. Safa’s anthology ganj e soxan (Teheran 1961): Vol. I: 23 11 59 (Bu Sakur) 30 5-6 71 (Daqlqi) 56 5-9 140 (RabeSe) 70 2 103 (Ferdousi 21 ) 165 8 68 (Asadi) 209 5-6 111 (Azraqi) Vol. II: 97 3 82 (Sana i) In addition to the above quotations more than a hundred single words have been quoted from sdhname in Chapter XII. together with references to the verses where they occur. 21 This be it is not from sdhname. PA RT TWO THE RHYTHM ; or rhetoric, he could not ope His mouth but out there flew a trope; And when he happened to break off I’ th’ middle of his speech, or cough, H' had hard words, ready to show why And tell what rules he did it by: Else, when with the greatest art he spoke. You’ d think he talked like other folk; For all a rhetorician's rules l each nothing but to name his tools, Samuel Butler XV HOW TO ASCERTAIN THE RHYTHM OF A GIVEN POEM §147 In a way this work could end here. The student who has mastered the contents of Part One can, with the help of the rules given there, ascertain the rhythm of any Classical Persian poem. There are, it is true, a few additional rules which he has to know, but they could easily be deduced without guidance 1 . Those who study prosody in order that it may help them to read and understand the poems correctly, but do not care about classifying and naming the various rhythms, and consider it unessential to know which rhythms arc preferred by which poets and for what genre, need not read beyond § 172. We shall now show how one can find out the rhythm of any poem or fragment of a poem which one may come across. * B Consider the following lines by j Farroxi ye Sistani: i J* ( it » a 1 ! * ** C* * «* •• C/ Sc J. J ')/£) if) )j hf 3 (/ m i * fXjs/t seed'd J 1 These rules are given below. See §§ 150, 152, and 156-158. 90 THh RHYTHM §147 Looking at the first mesra V we find that it may be scanned in the following ways: cielam dar jombes bar l - digar Cf. §§44-45 and 54, "My heart has once more become restless (lit. come into movement)". It is seen that the first mesra V unambiguously shows the quality of all syllables except the fifth and the ninth. The quality of these two syllables may be found by comparison with the rest of the poem 2 3 . The second mesra V can be scanned in the following ways: jT ’-P j fjf naddnam td - darad bdz ’ dar sar ce Cf. §§62 and 65 _ _ V _ "1 do not know, what it (o: the heart) again has in mind' It is now' clear that the ninth syllable is short. For the fifth syllable we need further material. g* ¥■ * hamdnd f andar pis ' darad Cf. §§89 and 90. Sesqt w ~~ ~ — — — ~ — "Surely it has love ahead". The fifth syllable is still ambiguous. bald i x' dhad f lu)r( ^ JN be man bar Cf. §§44 and 45. avordan "It will bring calamity upon me". Still no result for the fifth syllable. begardad td kojd binad be giti — v — — — w "It roams till it sees somewhere in the world ...” 2 Even if we possessed only the first mesra ^ of this poem, it would still be possible to Find out the rhythm. See below §§159-168, 3 In the literary tradition this word is often pronounced avardan. However, awrdan is not only common in the colloquial language, but also etymologically correct. §147 HOW TO ASCERTAIN THE RHYTHM 9 ! The rhythm is now clear 4 5 . Keeping in mind the definition of the rokn as a recurring section of the rhythmical pattern \ we may subdivide the rhythm as follows w ^ v_. — . It remains only to be seen whether the rest of the poem does follow this rhythm. For the next three ahyat all goes well: az /" sux i balaju i setamgar bar it mehr a- rad o bird" ha rad pdk mara az fa- mes b az x"a- b n az x"ar ze delha mar- donut" ra xei- mara ban ze del bit sad * kojd ydbam del i andar- r' basad home sar 6 x pr e x 'is "... a tyrant, someone terribly capricious (lit. more capricious than this), and always in trouble (lit. misfortune-seeker). On him it bestows its love and removes me completely from rest and sleep and food. From their hearts people receive good (lit. there is good to people); I for my part receive bad from the heart. Where can I get a heart worthy of myself?" The next mesrd^ apparently does not fit into the rhythm: del e saves - te ke forusad be gouhar “Even if the sword remains long (lit. much) in the scabbard ...” Here the last rokn is -- instead of ^ It cannot possibly be read otherwise. It now seems that the rhythm of the poem is - ^ — V V nasavad kon- d o nagardad honar e ti- q' nehdn "... it does not become blunt, and the sword's skill does not disappear". vur ce az ees- nr' nehd n gar- dad mdh an - dar m\q * 1 2 ■ j ■ ■ w w “And even if the moon is hidden from the eye in a cloud We are now forced to change the third rokn from ^ ^ — to The rhythm now appears to be V v -- — v_^_ nasavad fi- re wo afrit - x 'te basad be mian "... it does not become dark, and (i.e. but) is shining in the middle (of the cloud)", 10 The prosodists do quote such verses, but they are never met with in literature proper. See for example al-mo^jam, p. 134* (Ed. M* Qazvim, London 1909; hereafter referred to as aUmo Sjam . ) 1 1 At this stage of the analysis a long ninth syllable cannot be altogether excluded. 12 Note the additional rhyme andur miq i andar tiq {cf. § 136 k The rhyme arrangement of this remarkable poem is xa xa ba ba xa xa §152 HOW TO ASCERTAIN THE RHYTHM 95 Sir' ham si - r’ box ad gar- ce be zattji- r' boy ad w 'The lion. too. is a lion, even if it is in chains"; nabarad ban- d o qe/dde Saraf e si- >■ e zidn ^ v o “ “ v_/ w wW” "Collar and chain do not take away the fierce lion's honour". bdz' ham bd- r ’ bovad gar- ce ke u has - te bovad J v_> u V " “The falcon, too, «r a falcon, even if he is bound"; saraf e ha- zi az bd- z ' fegandan natavdn w \J w w “ T — u o - "The honour of falconhood one cannot cast away from the falcon". Not surprisingly, we now have to modify the second rokn as well. The rhythm of this qet*ie which is treated in §§211-212 is then seen to be ^ V _ _ Vf w V. V V t In the malfu: "pronounced” version (cf. §61) a long substituted for two shorts is by some changed back into (wo shorts. Thus, the mesraS var ic a: eei- nr' ncfid" gar • dad mdh an- dar miq may in the malfu: version become cf’b) •« f -v varce azces monehdgar datiomafum daromlq §152 The above poem teaches the following important rule; One long may always be substituted lor two shorts 53 , except at the beginning of a mesraS where instead a long may take the place of the first short. becoming This formulation of the rule is not quite exact: If the mesrd^ begins with then - „ — may be substituted for this, hut if the rnwru? begins with then ^ may be substituted for it. In other words, we have to distinguish the opening - from the opening “ v Of these the former is by far the commonest. Examples of the latter are found in §§154, 172. and 248. 13 The substitution of - for « « is particularly common in the last rokn oT a mesraS. Indeed, here - is the rule and is the exception. 96 FHE RHYTHM Si 52 It is important to notice that this is a "one-way-traffic rule", i.e. (and initially - w) may be freely substituted for but ^ can never be substituted for - (or ^). In other words, if in a given poem ^ ^ alternates with - (or initially with - ^), then w ^ must be considered the basic rhythm. Hence* it is not possible to scan the second tucsra^ of (he sixth heir in § 147 as del e sa ves- re ke farusad he gauhur W ” s/ y u ‘ ‘ y ‘ for the basic rhythm is ^ . One could, of course, argue tha! the basic rhythm is uw hut this would be contrary to the rule given in §150 [x three short syllables are not allowed together). §153 In the following lines ^jJL^o Amir Xosrov e Dehlavi lists the five /namaw-poems which make up his xamse “collection of The first mesraS may be scanned in the following ways: dadi aval dadi avval e davdr e davvar Cf. §§90, 1 10. and 54 v/ s./ "First, you 14 gave to the revolving vault ..." The next mesra V apparently removes all ambiguities and makes the rhythm clear: 14 The poet is addressing himself. HOW TO ASCERTAIN THE RHYTHM 97 musand i zc mat la % l-'anvdr W ~ — ^ — YJ “ — "... light from The Rising-place of Lights' 15 ". But the next mesrdS shows a variation: e kardi afgah ' bd nasal \ tamam \y — — o ” M “Then with complete joy you poured (lit. did) ..." The end of this mesrd V cannot be two longs, but as we saw in the previous paragraph, one long may be substituted for two shorts. Wc therefore read the ezdfe short and arrive at the following basic rhythm: The rest of the abydt quoted follow this rhythm: sahd e sir in o xosrov andar jam bdz ■’ dar $ dlam e xeradmandi sur e majnim o leili 1 b afkandi pas zabd" por dor e dan kardi sarh e rdz e sekandari kardi v i" zamd " k az java her e anjom minegdri sahife ye panjom "... the honey of “Sirin and Xosrou ’ 17 into the cup. Again, you threw the agitation of ‘Majnun and Leila ' 18 into the world of letters (lit. of wisdom). Then you filled your mouth with pearls of Persian and explained the Alexandrian Secret 19 . And at this time when you arc writing the fifth volume with jewels of stars ..." Here follows a detailed description of the contents, after which Amir Xosrou adds: pas nevisam ze kelk e mosk 'seres t ndm e i" hast ' xane hast ' behest ■ ™ w U — 15 See §218. 16 Cf. § 127. 17 See §195. (8 See §253. 19 Amir Xosrou here alludes to his masnavi IJl Hint' ye sekamktri “The Alexandrian Mirror", see §180. 98 THE RHYTHM $153 “Then lei me write with the musk-mixed pen (o: pen dipped in ink) the name of these eight houses (o: chapters): The Eight Heavens’’ 0 ". In §164 below, it will become clear that the rhythm is realty: S his would also have become clear, had we read the continuation of the abyat quoted, which as the reader will have guessed are taken from the exordium of hast behest. It is not easy to sec how this rhythm should be subdivided into arkan. This problem will be dealt with in §§226-230 where this and related rhythms are described. § 154 As a final illustration of the technique of ascertaining a rhythm we take t \azal no. 6 from di van e hdfez (ed. Qazvini). The matlaS “opening verse” is as follows: be moldzemdn L - solid" resdnad [. do % ra • e ke i" Cf. §§ 54. 65, and 44-45 % * Who will deliver this request to the attendants of the sultan": ke be sokr e pddesdhi ze nazar mat'd" gedd rd "(that) in gratefulness of (your) kingship, don’t drive away the beggar from (your) sight". Here the pattern has become clear already from the first beit. A perusal of the rest of this qazal will fail to show any variation from this rhythm, which is described in §248. It is a characteristic quality of Hafez’ poems that it is very easy to find out their rhythms —another instance of ravani “fluency" — indeed, the rhythm could have been deduced from the first mesret ? alone with a fair degree of certainty by bearing in mind the information given in §57 (e commoner than e), §76 (- e very rare, not used by Hafez at all), and §51 (in commoner —and in Hafez much commoner — than in). By applying the method to be explained in §§159-168 below. 20 See §23 2, HOW TO ASCERTAIN THE RHYTHM 99 § 155 the rhythm could have been deduced from the first mesrd^ with absolute certainty. §155 In practice, ascertaining the rhythm is not always so easy as the above examples may lead one to think. It has been assumed that the reader knows from the outset the correct interpretation, i.e., whether to have ezdje or not, whether to interpret .xJi as sokr "gratefulness' or sakar {or sakkar ) “sugar " 31 , whether to understand ol*j as be moldzemdn "to the servants" or be mol la zamdn “to the mullah time If one is not able * to decide upon the proper interpretation at once, one has to reckon with all the various possibilities while trying to determine the rhythm. When comparison with the following abvdt has made the rhythm clear, one may then be able to eliminate a number of wrong interpretations. Let us imagine that we have decided to read Nezami's xeradndme "Book of Wisdom" or — as it is sometimes called — a-LJUI eqbdlndme “Book of Good Fortune” 22 . The poem opens with the following be it : The first half of the first mesial is made up of the following letters: xrd hr kj\ If the printing is not careful — it rarely is — we will not be able to tell where d belongs: xrd hr or xr dhr or perhaps xrdhr. There cannot be much doubt that kj means kojd “where”, but to the beginner xr-d~hr offers a number of bewildering possibilities: xord bar kojd “the little one everywhere", xarad bar kojd “he buys everywhere”, xerad e bar kojd “the wisdom of everywhere”, xerad bar kojd “wisdom everywhere”, xar e dahr kojd “where (is) the ass of the age?”, etc. etc. The corresponding section of the second mesra V reads: c n'm xd\ It is difficult to make anything out of this except ze nant e - xodd “from the name of God” v-. This immediately eliminates most of the above alternatives and leaves us with xarad bar kojd and xerad har kojd. Which of these should be preferred must be inferred from the context. In the end even the beginner will probably arrive at the following result: 21 A number of such misreadings are shown in Acta Orientalia J9 m page 242, Copenhagen 1978, 22 xeradname is the second half of Nezamfs eskandarndme “Book of Alexander". The first hall ■ ■ if is called J, sarafname "Book of Glory", 100 THE RHYTHM xerad bar kojci gan~ j i brad padid ze ndm e xodd sd~ zad d“ ra kelki “Wherever wisdom finds a treasure, it makes of God s name its key ' 23 . This bed is not very difficult, but one does come across baffling poems, which apparently have neither rhythm nor meaning. However, even the most difficult (/aside has some lines which are less difficult. So, if one cannot make anything out of the first lines, one should glance over the following lines, find some which one understands better and ascertain the rhythm on the basis of those lines. The knowledge of the rhythm may then enable one to read and interpret the first lines correctly. §156 One more important rule of Classical Persian prosody should be carefully noted: The sequence may be substituted for the sequence as in the following be it by ^ y Sa^ide bne Abe 1-Xeir: go f tar e deraz' moxtasar bay ad kard v az yar e badamuz' hazar bay ad kard — VJ v j 24 “Long speech one should make short, and from a friend teaching evil one should keep away". The alternation of - uu and w- is very common in the middle of a robd^i {see §§255 foil.) and extremely rare elsewhere, but it does occur as may be seen in the following heit by jjj Rudaki; ## n iz ’ aba 2 5 nikitd" nemayadat jang ' jand laskar e f dryad ' ni x "as'le ni sud 'mand _ _ 23 For the sense cf. §219. 24 The basic rhythm is 25 See §121. V.V , sec 5 fol. HOW TO ASCERTAIN THE RHYTHM 101 $158 “Further, with beauties vain words 20 wage war for you. The army of lamentation (is) neither wanted nor useful" 27 . Instances of — — other than robust may also be seen in $§236 and 260. § 157 Very rarely an overlong syllable is found to coexist with a long one in the mathematical middle of a mesra V. Such cases will be dealt with under their respective metres. See §§ 186*187, 223, 225, and 240. $158 In pre-classica! poetry a short syllable is sometimes elided or added at the beginning of a mesra *1. and other irregularities may also be met with. ** ^ X ** ama* dam ke~m 2H x'*u$ *far' nisi yd xahr e mum ioz in ydr h nisi p +» i- ■■ (‘>1 \J~Zj w v “Know thus, that 1 have no lover, or (even) have no friend in my country but you '. >£ 3 j - 9 - Xosravi This rhythm is dealt with in §242. In classical poetry these irregularites are not met with and hence need not occupy us here. 26 Can also be translated “lie" or "deceit". 27 Cf. Llwell-Sutton, p. 125. 28 Or k-am Cf. §§79 and 101. XVI THE CIRCLES OF THE PROSODISTS § 159 In order to ascertain the metre in the way described in the previous chapter one usually needs several ahyat. However, employing an ingenious device, known as oJ\i da ere "circle", it is possible to ascertain the metre of a given poem on the basis on a single mesrd < i. The overwhelming majority of rhythms employed in Classical Persian poetry can be divided into five groups. The rhythms of each group have so many rhythmical features in common that they can all be represented as different sections of one "circle". §160 More than 80 per cent of the metres ordinarily met with in the literature can be fitted directly into one or other of these ’‘circles", and all the metres which do not fit belong to the least commonly employed, so that actually about 99 per cent of all Classical Persian poems fit directly into the circles. Moreover, with just one exception 1 2 3 , the remaining metres - may all be fined into the circles by the addition of just one short syllable or one short vowel in the (mathematical) middle of the rhythm §161 The five circles: I: II: 1 Namely - see §§241-243, 2 o: of those ordinarily met with. 3 See §§ 184-188, 223. 235, and 240. THE CIRCLES Of THE PROSODISTS 103 $163 V: The “circles" should be read clock- wise, beginning with the starting point at “twelve o'clock". §162 Circles I, II and IV correspond to the fourth, first, and second circles of af-mo^jam (p. 71-72). Circles I and H were taken over from the Arabs; circle IV was discovered by Sams e Qeis, the author of al-moHjam, himself. Circle HI was discovered thirty years later by J?** Nasir od-Din e Tusi in A.D. 1251. Circle V is nothing but circle IV read counter-clockwise*. §163 Writers on prosody, ancient and modern, have invented a number of circles to account for the remaining metres 4 5 6 7 . Llwell-Suiton. too. who prefers to call his circles “tapes or ribbons of indefinite length” has added no fewer than nine circles to account for the remaining “0.1 of the metres actually used by the poets”'’, most of the additional circles being identical to circles found in Xanlari's earlier work on the same subject T . 4 Cf. Elwell-Sutton, pp. 75-79. 5 One of these circles is given in §243. 6 ElwelE-Sutton, pages 86-87. 7 ojj torn e SeSr e far si, Tehran 1958. 104 THE RHYTHM $163 However, instead oi adding to the number of circles, il is better to classify those very few poems which do not fit into the above five circles as exceptional rhythms. Those poems do not testify to the existence in the poetic tradition of rhythmical patterns other than those shown in the five circles above (with the sole exception of the rhythm treated m §§241-243); they only show 1 that the individual poet is always free to choose any rhythm he wishes. He may compose a poem in a Thythm which no one else has ever used and which does not lit mlo any of the additional circles of the prosodists. An instance of this is given below. § 172. § 164 A couple of examples will show what is meant by "fitting the rhythms" into the circles: bahr e motaqdreb (§27) may be obtained by counting eleven syllables forward from the starting point — "twelve o'clock"— of circle I: ^ — <-» - - \j “ w “ i By counting twelve syllables from the same point we get bahr e motaqdreb e sdlem (§ 73) : ^ — w — ^ u — . The rhythm of the examples quoted in §§ 147 and 82 second example may be obtained by counting eleven syllables from the starting point of Circle II: w ‘ . If we begin our counting w ith the fourth syllabic after the starting point of Circle II — "half past two" — and count eleven syllables forward we get the rhythm of the examples quoted in §§100, 112 last example, 126 second example, and 127 last example: v v-. If we count fifteen syllables we get the rhythm of §111 last example: ” V W — ^ \ If we count fifteen syllables forward from the starting point of circle III, we obtain the rhythm of the example quoted in §148: w w — ^ ^ - w ^ — w w It should be noted that the circle shows only the basic rhythm, not the variations noted in §152: ^ - ^ - ^ The rhythm of the example quoted in §153: — - does not fit exactly into any of the circles, but if we start our counting with the fifth syllable of circle IV — "a quarter past three” — we obtain a very similar rhythm : Remembering the rule given in § 152, we take this as the basic rhythm and conclude that this masnavi follows the rhythm: w yu ^ w — V ~ “ * Counting sixteen syllables forward beginning with the third syllable of circle V — "a quarter past two" we get the rhythm of the example given in § 154: §165 On the basis of these circles. Eiwell-Sutlon prefers to give the rhythms “code-numbers". § 166 THE CIRCLES Oh THE PROSODtSTS 105 rather lhan employing the traditional terms. Thus, the rhythms shown in the previous paragraph arc labelled respectively: 1.1.11, t. LI 2, 2.1 I I, 2.4.1 ], 2 4. 15, 3.1.15, 4.5 . 1 L 5*3,16. 5.3.16 means that you should take circle number /hr, begin your counting with syllable numhcr threw and count sixteen syllables. This method is quite ingenious, but unfortunately Elwell-Sutton tries to squeeze all possible rhythms into the circles, by allowing certain sections of the circles to be jumped over or repeated. Thus we arc told to accept that v - is a variety of circle V, and that v m v *- ■ w - - - vw is a variety of circle IV H . T he "code-numbers" of Elwdl-Sutton will be given together with the traditional terminology in Appendix Gnc P CL 5$ 186 end and 245. § 166 If one has a short poem or a fragment of a poem too short to allow the rhythm to be ascertained by the method described in §§147-155, one can practically always find out the rhythm by comparing with the above circles. Here follow a few r mesra 5' of the type one is likely to read or hear quoted out of context. ^ This may be scanned in the following ways: gamiom a ~ gandom J ou :e J ou Cf. s§ 44-45, and 97. ‘'Wheat grows from wheat, barley from barley” Moulavi This cannot fit anywhere but in circle II — begin with fourth syllable - the rhythm must therefore be: - ^ This rhythm is described in §§201-203. v, ^ ^ #♦ har^ez namirad P_ de/as ( n ^t. sod be ^esq Cf. §§44-45, 64 and 65. ' a" ke zende * _ V _ “Never dies he whose heart has become alive by love" Hafez This can only fit into circle IV— begin with the seventh syllable — so the rhythm must be: — — w - ^ which is described in §§237- 238. 8 Elwell-Sutlon. pages 87-89. 113 and 188. 106 THE RHYTHM $166 kosa* manegar ° e tnangar o ehsan c - x*ts e Cf. §§ 54 and 102. KJ V/ ^ ““ “ v “ "Do not look at others’ faults and at your own goodness". Nezami Since three shorts cannot occur together y , this may he simplified to — w- — Except for the last syllable this fits with circle III counting eleven syllables forward from the fourth syllable: - ww — Re- membering the rule given in $41 that the last syllable in a mesra V is anceps, i.e. indifferent with regard to length, we interpret this section of circle III as-wv — w v-- v- This fits with the above rhythm, which is described in §216-219. When comparing a mesraS with the circles in order to ascertain the rhythm one should take the most ravan "fluent " version as one's point of departure. Thus, in the first example above, since a: is more ravan than 'a: (§51) and heruyad is more ravan than heruyad (§97), one should first try to fit gandom az gandom heruyad jou ze jou - ^ — - w Similarly in the second example where an. ke and zende are more ravan than 'an. ke and zende respectively (§§51 and 76) one should first try out hargez namirad an ke delas zende sod be esq - - ^ m - v- v In the third example, since e and a are more ravan than e and o (§57), one should accordingly first test V eih e kasa 11 mangar a ehsan e X"ii In all three examples the most ravan versions happen to be the rhythmical ones, but when this is not the case one has to try out the permitted rhythmical variations one by one. This job is less difficult if one follows an order of increasing sangini "heaviness”, cf, §76 end. § 167 It does happen, though rarely, that a line may be interpreted rhythmically in more than one way. Consider the following beit by CJL 1 Axavan e Sales. * i* AS 1 L Jars - tar Lx a be man mozde e ja-baxS i dad Cf. 54. "The lesson of history gave joyful news to me: ..." 9 See § 1 50. 10 Cannot possibly be scanned mozde. see §70. §169 THE CIRCLES OF THE PROSODISTS 107 zur az bazu ye sarmaye , , - , . , . . 7 . S - be dar x" ahad sod zur ’ az bazu ye sarmaye ■ Cf. §§45, 54, 64, 97. ^ ^ V ^ ^ "... ‘Power shall go away from ihe arms of the capital*, (o: the capitalists will lose their power)”. Comparing this with the five circles, we find that two interpretations are possible: either: dars e tdrix' be man mozde ye jd’baxs / dad zur' az bazu ve sarmaye he dar .v "ahad sod n v * ■ \J ~~~ ~ vy W — \J w or: dars e tdrix' be man mozde ye jd'baxs i dad zur az bazu ve sarmaye be dar x "ahad sod ■ «> wr ■ * The first alternative fits into circle 111 beginning with the first syllable: yhe second alternative also fits into circle III, but this time beginning with the fourth syllable - v v-- However, the line zur az bazu ve sarmaye be dar x' dhad sod & a v * ■■ is very heavy, so it is a fair guess, that the first alternative should be preferred. § 168 The reader may then look up these two rhythms in Appendix One and check with the references given there. This will show that the rhythm ^ ^ - - w- with its variants is very common, whereas -wv- -wv, — with its variants is a very rare one. This confirms our impression that the first rhythm is the correct one, but we cannot be absolutely sure till we find the poem from which the quotation is taken 1 1 . §169 Now consider the following beit by j(JL- Salman e Saveji: II The poem can be found in Axavan e Sales: jy*,! arqanun (Tehran 1345-48). p. 32. 108 THE RHYTHM $169 . , e to harm ve . ... x at v to , e ... , , lab - - , - - • - loSloh . . markaz _ laic e to hanv ve xatt e to e ■ Cf. $$ 54, 74, 90, 1 1 0. S' S' w w V V "Your lip (is) the protector of pearls (3: teeth), your down the centre of tulips (o; red cheeks)". The number of possibilities is bewildering and only slightly reduced by comparison with the second mesra 5' : sab l - home / koukab mah _ ll l ha \ at hale 1 2 e to e e to c Cf. $$ 54,74. w V V _ u "Your night is carrying stars, your moon has the halo line” 1 3 This rhythm evidently fits into circle II: ^ ^ ^ ^ , but it also tits into circle III: w ^ — w w - - . and circle IV: ^ ^ — . As a matter of fact, this belt was composed on purpose so as to fit into several metres, in order that the poet might display his dexterity 14 . These three rhythms are treated in $$191-193, 210, and 227. § 170 A poem with a double rhythm, one which can be scanned throughout in two ways, is called JiJ^ S zo l-bahrein "having two metres" 1 5 . In most cases such poems sacrifice content for form with disagreeable results, but there are exceptions. The poet, ^Ul Ahli ye Slrazi, has a mystical masnavi, 540 abyat long, called J^L>- ■ se br e hafal "Lawful Magic". Not only is it zo l-bahrein and abounds in clever wordplays, but it also has passages of great beauty from the point of view of content. 12 See §41. 13 Wordplays make this mesra V untransla table. The night signifies the black hair; by star the lustre of the hair is intended, i he moon which shines in the night is the beautiful face surrounded by black hair, xatt apart from the meaning ‘line"' also means down of the upper lip or ihe cheeks (as in the first mesra V), and since a halo round ihc moon is a bad omen, hale has acquired (he meanings: "author of evil', "destroyer", "corrupter"* '"seducer", so that the implied sense of the mesra G becomes something like: "The down on the checks of your beautiful face framed by your glossy black hair is seducing". 14 Cf J Homai, op. cil. (§129 above), pages 80-81, 15 The example in § 169 ought to be termed y* zo l-hofmr "having several metres". § 1 70 THE C IRCLES OF THE PROSODISTS 109 S' *?(J bf, w J *• * •*— • pis 'tar ( }az marg e x'qd ( )ei x'dje mir Cf. §$44-45. id savi az marg e x"qd (')ei x'dje mir Cf. §§44-45. and 90. a "aje ilar (')ab- risom d mu m dar gelim Cf. §§44-45, and 54. : late) never (lit, when?) intends to show pity (qamx'dri). Its kindness following upon affliction (qam) is ignominy 16 (x'dri). God has embellished your form like (cu*) a fairy (pari) that you may fly (part) towards the matchless (bicu") Creator. Pass on from this form and its appearance (hinaias). Do not consider the sound of the reed-flute, but (lit. and) look at (bin) its Player 1 " ? (ndias)". 16 It is ignominy, because by accepting the kindness of fate to ourselves we tacitly accept its unkindness to others, 17 For (he sense of the last hat, compare with the hat quoted in $ tOO. no THE RHYTHM §170 These two rhythms are treated in §§201-203 and 216-219. The student is advised lo study these ahyat carefully. They will serve as a revision exercise, the more important rules of Part One being illustrated in them. Moreover, their beauty is of a type which cannot be comprehended either on the first or on the second reading, so that a careful study will also be rewarding from this point of view. § 171 In the following chapters all rhythms which are more or less commonly met with in Classical Persian literature have been described. It would be vain to aim at completeness. Even if one could read all the extant literature and describe all the occurring rhythms, a manuscript might yet turn up containing a qazal in a hitherto unseen rhythm, or a traditionalist might write one. For the poets arc not restrained by the rules of the prosodists. They are free to create new rhythms and use them. § 172 The two abyat below are taken from a mystical masnavi by Seix Bahai called j sir o sakar “Milk and Sugar”. This masnavi comprises 150 abyat and is written in a metre which 1 have not come across elsewhere in Classical Persian literature. It cannot be fitted into any of the five circles 18 , and moreover, although the mesra 9 of a masnavi ordinarily comprises eleven syllables, in this one we find twelve syllables (cf. §139). The rhythm lSvjrv/“^ WU Yy As we know - may be substituted for ^ v even in the first rokn (§ 1 52). In the example below we see that in one mesra $ is nowhere substituted for in one mesra 5 - is substituted once, in one mesra V it is substituted thrice, and in the last mesra - has been substituted for ww throughout: bj #4 * ** 1 ) yd rah be kard- m be heda 4 - w v — yat e pi- \J -*■ mi git ^ — bd zou- } {JU mat e ah- I e vafa u v *“ w s*rava- n c vafa v u “ \j — q o del e dgdh v v — — — 18 It does not fit into any of ElwelbSutton's additional circles either, but cf, Xanlan, op, eiL (§163 above), p, 211. §172 EXERCISE 111 ft a l ki- lt a l ki- lt a (liih "Oh Lord! With the generosity of the faithful, (and) under the guidance of the leaders of Fidelity, keep saying with joy 1 g and a vigilant (or knowing) heart: ‘Allah! Allah! Allah! Allah!' " For the traditional classification of this metre, see §209. EXERCISE: Determine the rhythm of the following bat by j U* Jami, and henceforth never read a line of Classical Persian poetry without ascertaining its rhythm. "The glory of Mahmud is gone and nothing that he did not recognise Ferdousfs worth". * * U*'b/>S has remained in time except ihis anecdote 19 According to Stein gass zouq also means “distinction of truth and falsehood by lhe light of divine grace"* XVII METRES WITH THREE SYLLABLES TO THE ROKN (THE FIRST CIRCLE) §173 bain: e motaqareb “the nearing metre" According to the traditionalists the basic form of this metre is fashion - - repeated four times in each mesrd k i.e. eight times in a bed, thus: # • * At* * \j (J 'j>/ s ' V As (J it* l V cf f* c/ ^ hami td ■ jaha" mar be cehre ■ .i be afkd- konad pi- fa fa rd so dan cu" /' mono ri- se ‘iudat m to mar sa- por i kei de sou mar hami kon beri rd tax dni pari rd xj “ " w — u ” ' \J “When (it) makes (it) a custom, make (it) a habit; (that is when) the world (makes) injustice (its custom, make) thou patience (thy habit). When canst thou become like a fairy in face? Become like a fairy in (thy) actions"! Naser Xosrou §174 This metre is not very common. Examples of it may be found in Gt 30. 48. 102. 182, 262. 288: Oil 250, 283: OH I 273; H 492; S 577 (T -at/, bena:). 694 (T -ant. agar). 816 (X -atigi ) ; M 494. 961-962. 1289. 2087-2088. 3116-3127, 3198. (These abbreviations are explained in the following paragraph.) Another example was given in §71. See also §229 footnote 5. FIRST CIRCLE: MOTAQAREB 113 $176 §175 For each metre a number of references for further study is given For this purpose the most widespread anthology of Persian poetry and the three most popular divans ' have been used. The anthology is Z Saf3\ \>tmj e soxan 'The treasure of Words' (-ill, Tehran 1961 ’ abbreviated GL Gil, GI1L references being to the pages. The three divans are those of Hafez, Moulavi and SaSdi. The standard editions of Hafez and Moulavi have been used: Foruzanfar's edition of Moulavi called divan v kahir 'The Great Divan" FX, Tehran AIL 1336-46 -abbreviated M, references being to serial numbers— and divan c hdfez edited by Qazvini and QanL Tehran 1941 - abbreviated H, references being to serial numbers. So far no edition of SaSdi has been produced that deserves to be called a standard edition. I have used i s j ^ koHiut e sa*idi '"Complete Works of SaSdP edited by Foruqi ASlianL Tehran A.H, 1354 -abbreviated S, references being to pages Being aware, however, that many may not have access to this edition, 1 have after each reference indicated the section: 0 iqtis&\'d c ! ani "Persian Odes")* Mr marasf "Threnodies")* Ml (oUi* molamma^dt '’Macaronic (Bilingual) Ones"), T ( OUt ray what "Fine Ones' 1 ), B ( haduve^ "Cunning Ones"). X ( xavutim "Gems"). Qq I v qadim "Old Sonnets "), H (oil/ hazaUdt "Obscene Ones"); after this the rhyming clement 2 has been given, and where this is not sufficient to identify the poem* 1 . 1 have also indicated the first word(s). References in italics refer to masnaviat. Apart from this the most important mamaviai arc mentioned expressly under their respective metres. For further information about these the reader is referred to ian Rypka: History of Iranian Literature. Dordrecht 1968, Imnisehe L it era turgesch ich le , Leipzig 1959 abbreviated Rypka. references being to pages (of the English and German versions before and after the shilling stroke respectively). §176 As explained in §73, the above rhythm is called bahr e motaqdreb e mosamman e sdlem “the sound eightfold nearing metre", generally referred to simply as bahr e motaqdreb e sdlem. Other rhyt hms, in which the basic dement is still the rokn , w — fa^ulon, are considered to be variations of the same bahr ’‘metre" (lit. “sea", note the plural j ^ bohur), thus sdhname is in the same bahr as the above ahydt by Naser Xosrou, but in a different xazn “rhythm" (lit. “weight", note the plural jljj! ouzdn ), being no longer sdlem “sound", but mahzuf “cut off' or maqsur “shortened". Cf. §§27 and 42. However, the traditionalists group together in one bahr “metre" not only rhythms which are strikingly similar as e.g. bahr e motaqdreb e sdlem (w-- v ,__ v — ) an d bahr e moiaqdreb e mahzuf (^ — ^ » — w-), but also rhythms which have hardly anything in common. 1'hus the rhythms of Circle III and even Circle V are taken to be variations of the rhythms 1 The meaning and’ arrangement of divan is explained in §§264-267. 2 l.e. the identical portion, cf. § 130. 3 The procedure is explained in §§ 264-265. 114 THE RHYTHM § 176 of Circle II. E.g. the rhythms — (Circle III) and --w-v— w-- (Circle V) are classified as varieties of gj* hahr e hazaj "the shaking metre", of which the basic form is ^ ^ w In our classification we shall not follow this grouping of basically different rhythms into one and the same bahr. Nevertheless, for every rhythm treated the traditional name will be given and it will be explained by what mechanism the rhythm concerned is thought to be derived from the sdlem “sound" one. §177 Now consider ihe following two abyat taken from Hafez's sole qnzai in this vazn namihi- ■ ilelam YU if raflqd* 1 ke gui mini az ham- sod a i qos- amir* mihudas llama** h(- sp sdqi d e sohhat r' A \"od <5- r 1 har jar kojdi sekastand s'ndi "I do not see any of the friends (being) true, My heart has become heavy (lit. blood)' with grief. Cupbearer! Where are you? The friends have broken the pledge of friendship, so that you would say, acquaintance itself has not existed '. It is seen that the last syllable of the first mesraS of both ahydf is overlong, As we have learnt in §41, this ts perfectly natural. However, ihe traditional prosod ist considers this a deviation from the standard rhythm, sekasfami is not described as faHuhm v - -* but as oiyi fakulan ^ , and the mexrdS is no longer sdlem "sound", but ^ — ■ moxbaq "made ample". The complete description of this rhythm with the traditional terminology becomes rather complicated. The last rokn in the first mesrd ¥ of a hvit is called Haruz (lit. "rhythm") 4 5 , The last rokn of the second mesrdk is called zurh (lit, "beat"). Hence the vazn "rhythm" of the above lines is described as jlL- > jC- hahr v motaqdreh e mosamman e moshaq ol-Saruz r a Salem oz-zurh. Hereafter the special terms denoting overlong varieties of the final arkdn will be ignored. They can however be (bund in Appendix One. 4 Cf §§77-78, 5 The sdme word occurs in the term V/jh e karuz "the science of rhythm, prosody". $179 FIRST CIRCLE: MOTAQAREB 115 § 178 By reducing the number of syllables of the standard metre to eleven —the number of syllables employed in masnavidt 6 — -we arrive at hahr e motitqdreb v mosamman e mcihzuj , ordinarily just hahr e motaqareb, the most well known of all Classical Persian metres. As this metre has been exhaustively described in Part One, no examples are needed here. Still I cannot refrain from quoting two abyat which — taken from one of the two most famous works written in this vazn, namely SaSdfs hip fan — contain a , tazmin “quotation" from the other one, Ferdousi's sdhndme : / ce x u os go/ - ke rahmai maxdzd- ke jd " dd- /•' Jerdou- bar a” lor r *’ mur / # rad d id- si ye pa- bat e po- ke ddne- n e stri k’zdd k ,f bad kas as/ x"as asi “How sweetly said the noble Ferdousi, (wherefore) may (God’s) mercy be upon that pure tomb (of his): ‘Don’t molest etc.’ (see §84)“. §179 This vazn is first of all the rhythm of epic poetry 6 7 8 9 , and it is almost unthinkable that a Persian poet should write an epic poem in any other vazn. though, curiously enough, the first attempt to write a idhndme was not in this vazn*. Ferdousi’s sdhtidme, which has a secure place in world literature, has had innumerable imitations. The most outstanding among them is zafarndme 'The Book of Victories" in which Mostouft in 75.000 abyat “ -one quarter longer than Ferdousi’s work - continues sdhndme down to the year 1332 A.D. Another able imitation is the much shorter **1^^*.; teimurndme or tamurndme "The Book of Tamerlane ’ by JU Hatefi Id. 1520), which comprises about 4.500 abyat. Less remarkable imitations with more high-sounding titles are the works by Ahmad e Tabrhri (contemporary with Mosloufi). ^y-li Qasemi (16th century) and 'Jt- pz* Fath SAli Saba (d. I ft 22). all three entitled idhansdhndme, Also zardtoitndme “The Book of Zoroaster" by py t Bahram e Paidu, the only Classical Persian 6 Sec §139. 7 Cf. Hermann Ethc: Neupersische Litteratur. pp. 229-255 (in Crundriss der iranischen Philologie //. Strasburg 1896-1904). See also Z. Safa : hamdsesardi dar Iran* Tehran 1324, and Ahmad SAli Ahmad: cd* tazkere ye haft dsmdn. Calcutta 1873, repr. Tehran 1965. 8 See §195. 9 One of these was given in § 128. 116 THE RHYTHM §179 work of note produced by a Zoroastrian, may be mentioned here. (Rypka 151-162 152-164, 165-166/168. 285 & 444 & 496/ 2 76. 326-327/313. 34/34.) §180 Another chef dceuvre of Persian literature written in buhr c motaqan b is vskumktniamc "The Book of Alexander", by Nezatni (d. 1209). This work is not. as its title might lead one to think, merely an epos describing the legendary exploits of Alexander, but what might be called a philosophic-epic tmsnavi'**. This work too has many imitations of which uUw vr sektmthtri "The Alexandrian Mirror" by Amir Xosrov e Dehlavi (d. 1325) approaches the original in literary merit. Also xcnulnamc it wkattdari "The Book of Alexandrian Wisdom" by Jami (d. 1492) must be mentioned here. (Rypka 212/204, 287/277.) §181 Another type of masnavi for which the tmfuqaivb metre is not infrequently used is the so-called "(moral) advice", and the fragments of t-L* afarinname "The Book of Praise (-worthy Conduct)’ 11 written in 944-948 by BO Sakur are in this rhythm. The most famous work in this category is Saadi's great work husian "The Fragrant Garden" 15 , completed in 1257 (A.H. 655). • v ♦ i J>/? ze svs sail ke port hr ftKw' hu - mh( h na- V d' panja- m inn da- ft o pant r* mmj "It was fifty-five more than six hundred, when this glorious treasury became full of pearls’ 1 . {Rypka 144/144, 250-253/241-245), § 1 82 Occasionally, especially by the earlier poets, masnaviat with romantic or mystical content have also been written in this vu:n, E.g. aUll? j ujj van/e wo golsah "Leaf and Resetting'" by ^Ayyuqi {10th century), yusof o zoteixa * 3 attributed to Fcrdousi, Oji \A $ homay o tw maxim composed in 1332 by X*aju ye Kerman!* fcrdqndme ‘'The Book of Separation" composed in 1360 by Salman e SavejK and iu* i * mohit e aSzam "The Great Ocean" written in 1681 by Jju Bidcl, (Rypka 177/175-176, 157- 158/158-159. 260/253. 262/254, 518/407.) Finally it must be mentioned that Hafez selected this rhythm, which he does noi employ elsewhere, for his short masnavi oLJL. sdqinamf “Song of the Cupbearer" This. too. has had many imitations, the prettiest of which is the saqiname by Zohuri (d. 1615). (Rypka 262 & 265/254 & 258, 30 1 & 725/29 1 . ) §183 Outside the musnaviul. i.e. in qasatnl. qazaliat etc. the motuqareb metre is not very common. References: Gl 18. 2! -24 . 35-40. 46 . 56. 70-9H. 127, 130, 164- 17 f, 196-204 . 271. 274. 298: G1I 13-17. 15H-160 . 1HL 218. 25 L 294: GIN 10. 1H1-1H3. 1S9. 257, 312. 329-331: S 599 (T -a ravu/uf), 601 (T us hud). 763 (B -an/ e man): M 239-240, 495. 1330, 2089-2090, 2251. 3128-3131, 3199. 10 Quotations in §§6(J. 87, 90. 97. 98, 102. 130, 133. 134. 136, 138. 1 1 Quoted in §59. 12 Quotations in §§69, 77. 94. 98. 99, 110. 126. 127. 145. 146. 13 The Persians do pronounce UJj in this way. <5185 FIRST CIRCLE: MOTAQAREB §184 The following abvat by SObeid e Zakani (d. 1371) are wriuen in a rhythm which is traditionally— and not without some justification —con- sidered a variety of the bahr e motaqdrch : mar a sckaval ■ ccsm c Sobeid ar * digar nabinad bar has ze das man a: tis 'nd / ■ siras nahinad * ccs mas bald i "Everyone has a complaint against an enemy; our complaint is against a friend. If the eyes of ^'Obeid do not look their fill at him, then his eyes do not see any misfortune (i.e. all other calamities are trifles beside the calamity of not seeing enough of one's beloved or he does not sec any other misfortune because he dies of grief)”. Remembering the definition of a rokn as a recurring section of a rhythmical pattern (§28). it is clear that each mesraS can be subdivided into — ^ — . This subdivision is by the traditionalists further analysed as — in the traditional terminology U*# faSlon faSulon . A glance at the above example shows that the traditional analysis with one exception (or d-s'na f) here agrees with the word divisions. As we shall see. other factors, too, speak in favour of this subdivision. §185 Traditionally then this vazn is described as fab Ion fa Sid on fa Sion faSulon faSbn fa Sul on fa Sion faSulon w w and faSlon — is said to be an Jj* aslam "castrated” variety of faSulon v — . Hence this vazn is called bahr e motaqareb e mosamnum e aslam "the castrated eightfold nearing metre”. One might object that this designation does not make it clear which of the ink an are aslant "castrated" Thai is true: an exhaustive description must specify in which ark an the modifications 1)8 THE RHYTHM $185 lake place, and in order lo understand such descriptions one has 10 know the traditional names of the several urkdth The first rokn in the first nn *srd*i of a heir is called sadr "breast, front part". Used in the dual, jtj+L* siuhein. this term indicates the first wkn of both mrvrdV 14 As we have seen (§ 1 77)_ the last rokn of the first and second mesrd S are termed Saruz "rhythm" and zarh "beat" respectively. The dual zarhtdn designates the last rnkn of both nwxraH* Whatever intervenes between smfrvin and zarhvin is termed ha\v "padding", the dual of which is jt j -!>■ hasavem. Thus: sadreiu huiv zarhem The complete description of this vazn is therefore hahr e moutqdrch e nmstimman e astern os-sadrem va sdlenr ra asiam oi-hasavein" r a salem oz-zarhein* However, as there is no vazn in ordinary use with which the above rhythm might be confused it is sufficient to say hahr e molaqdreh v a si am. § 186 This vazn does not fit directly into any of the circles. It is one of those ouzan which, as stated in §160, may be fitted into one of the circles — nn this case Circle I — by the addition of just one short syllable in the (mathematical) middle of the rhythm. Moreover an examination of this and similar rhythms lf * would reveal that the end of the second rokn always coincides with the end of a word and mostly with a break in the meaning, too. In other words, there is a pause or caesura after the second rokn, i.e. in the (mathematical) middle of the mesraS. If wc imagine that this pause has the length of a short syllable, the similarity of this vazn and of the basic form of hahr e motaqareh becomes striking: bahr e motaqareh e salem : ^ ^ v - - v - hahr e motaqareh e aslant Following Elwell-Sutton (p. 88), we term such ouzan "doubled metres” 1 7 and indicate the caesura (pause) by a vertical stroke: — W “ “ I — VJ " 14 The first rokn of the second mesrd C is called ehteda “beginning". Cf, §242. 15 Note that salem va as lam ol-hasavein means that the first and second rokn of the hasv are salem and asiam respectively. The order in which the terms are employed is important. (A mesra^ never contains more than four arkan . hence the hasv is never longer than two arkan). 16 See §§223. 235. and 240. 17 The traditional term is jJZ mokarrar “repeated". ^ 189 RRSTCIRCLF : MOTAQAREB 119 El well -Sutton's ” code-number"' is 1.2, 5(2), which means that we take Circle 0m\ begin with the wiont! sellable, count fire syllables and repeat nrnr. § 187 In doubled metres an overlong syllable may coexist with a long syllable before the pause in the middle of the mesial 1 8 , as in the following he it by Hafez. ... // * * M • '• * ^ hdfez xu* ha - ce nali vadat yard gar vas I _ _ dar ga- /’ .v "dhi h o bigdh “Hafez, why are you complaining? If you do want union (with the beloved) you must (be prepared to) suffer (lit. eat blood) in season and out of season « ■* One is tempted to say that in doubled metres each beit contains really four mesra 9, but this would be wrong for whereas overlong syllables are quite common at the end of a mesra £ in all ouzatu they are quite rare in the middle of a mesra *i in doubled metres. Elwell-SuUon gives statistics for the occurrences of overlong syllables before the cssura in the doubled metres Uza» mqfiaVdan) Jl ^>UiAi mofia Veld ton jIpU fdVe/on ( O^tc-U fdVelan) 0*}lp'i faVeldto ^y^U-U fd Veld ton ( j iJSU- U fa Veld tan or jLU-U fd Veil van) jJUi faVlon (O^U faVldri ) 1 9 J y*A* mafVulo 19 Often written jJ v to distinguish it from fa^ehm (Jordan). PI RST Cl RC Lt : MO TA DARLK 121 § 190 mostafSeh mostqfieion mostafSelan) mostq field ton ( jlJ — * mostaj^eiatan) mafSulon (oVjjuu mafSulan) mafiuiaio My first encounter with the afdSU was in 1967, when I attended lect tires on Ottoman poetry at the University of Ankara. These mysterious formulas exercised a strange fascination upon my mind and I felt that they were the Open Sesame of Oriental poetry. 1 have later learned that other orientalists* too, have fell a similar excitement when first confronted with the afa HU, and it appears from the poems of Mtndari that he* too, was fascinated by them. Enjoying their sound he not rarely uses them In his poems with a sort of dadaistic effect (see kollidt e sums e g. lines 299% and 34253), In one instance he gleefully mocks the afa HiL closing a poem with the line (22518) i/' cA* ^ mafuiSvhn fa 57 atafan mo/taWon faHfalafan mafia Het JaHlalan k e sirdzi he dust drad del e mdra he xdl e hin- day as ha x sum Samarkand b hoxara rd $ 193 SECOND CIRCLE: HAZAJ 123 "If that Shirazi Turk would take our (a: my) heart into his hand (eg. if he would make the least effort), 1 would give Samarqand and Bokhara for his black mole '! — if not his best, at least his best-known quzal we find that in a total of 72 arkdn. it happens only five times that the last syllable of the mkn does not coincide with the last syllable of a word. In this quud the use of internal rhyme, too. marks olT the arkdn. Thus in the second beii jw jtj hoick sdqi rhymes with mev e baqi “Give, O eup-bearcr the remaining 1 wine*\ and in the last he it qazai gofti rhymes with IYJ (for soffi 4 You have said (a: composed) a qazal ...” “and perforated a pearl" §193 This mr/i is quite common. It is the metre in which Sana ! (d. about 1 130) composed his famous qaside makon dar jex- "Rely neither upon body nor soul. For one is too low and one too high". References: Gl 128. 236. 265. 275; Gil 184. 198. 215. 264. 277; Gill 3. 31. 58. 89, 101. 103, 109. 121. 126. 163. [72. 174. 220. 283. 316; H 1. 3. 95. 115. 120. 121. 146. 149, 151. 153. 165. 194. 278. 288. 318. 327. 354. 356. 370. 374. 412. 440. 454, 474; S 447 (Q -am gardad). 511 (Ml -aqi). 581 (T -an dar ad). 609 (T -an dyad). 626 (T -ei. agar), 644 (T anam, agar). 645 (T -inam. re dastam), 657 (T -a ran, do cesm), 660 (T -//aw. xetdf), 669 (T -di. to a:). 669 (T -at. ve tuy). 676 (T -am kardi). 678 (T -andi. negara), 700 (T -ur bens ini). 706 (B -dr a, :c bad). 712 (B -i:at), 734 (B -an bd.iad). 735 (B -a namibaiad). 738 (B -d monad), 755 (B -aham), 761 (B -a ye durvisan), 770 (B -aw astj), 771 (B -ah ami), 781 (B -awi. bar dnatrt ). 781 (B -aw/, habar). 810 (X -aw oy). 811 (X -a /), 830 (Qq -ar mihinam). 831 (Qq -am numibimtm) ; M 54-72, 324-325. 562-593, 1023-1025. 1221-1225. 1293. 1314-1315, 1337-1340. 1412-1443. 1844-1860, 2118-2119. 2160-2169. 2290-2299. 2498-2562, 3180. i Or “(ever)lasting". 124 THE RHYTHM Si 94 § 194 By reducing the number of syllables to eleven we arrive at the ma.wcn7-variety of the bahr e hazaj. Cf. § 139. * I i y ce x "os hi meh- rahuni az do sat hi i ke vek sar meh * rahuni dar - d e sar hi ■*» - agar majnu" did e surf- de i dost del e lei Id 2 az u" surd detar hi a. - \j ^ ■ hj — — — w — ' "How lovely is love when it is mutual, for one-sided love is irksome. If Majnun had a desperately loving heart, the heart of Leila was even more desperately in love ". <1)1 y ^ Baba Tahcr e SOryan Each mesrak contains only three arkdn; hence each heit Contains six arkdn and the vazn is accordingly described as — « mosaddas "sixfold, hexametric'’. The last rokn is mahzuj The full description of the vazn is therefore bahr e hazaj e mosaddas e mahzuf , but the designation bahr e hazaj is sufficient if it is clear that the poem in question is a masnavi. See also the example in § 147. §195 The scanty fragments of the ninth-century sahname of ^sjj^ lPj* — • MasSudI ye Marvazi arc in this rhythm, but otherwise this is the rhythm of romantic love stories, the fashion being set vnth jyij j ^ ; j vis o ramin, composed around 1060 A.D. by J'&S Faxr od-Din e Gorgani. Earthly and heavenly love are often indistinguishable in Persian poetry. Therefore, this vazn soon became popular as a vehicle of mystical thought as well. (Rypka 1 152 / 152 , 177 - 1 79 / 1 75 - 1 78 .) Deservedly the most famous poem in this vazn is Nezami’s jtsr- s s s~*~ xosrev a Sirin. completed in 1175-76. As usual. Amir Xosrov e Dehlavi has an imitation almost as good as the original with the same subject and almost the same title: sirin o .x osrou. Jami’s best known masnavi. vusof o zotei.xa. completed in 1483, is in bahr e hazaj and so is the masnavi of the same name by jit LotlSali Azar (d. 1780 A.D.) as well as s J» gol o nouru: "Rose and Newyear " by X^aju ye Kermuni (d 1352). o-ij y ) jantihi o x u or.sid by Salman e Saveji [d. 1376], j »— > i sam'i o parvdne "The Candle and the 2 Or U’iii. SECOND CIRCLE; HAZAJ 125 $ 197 Moth” hy A hll yc Sir.lzf (d. 13X5). and the once so celebrated jU ) w ntuhntud o unw: by Zolall (d. 1615). (Rypka 210-213 202-205. 287/277. 260 253, 262/254. 286. 276. 301/291.) The prolific j'jae- xy Farid od-Din e SAttiir (d. about 1229) has a number of masnavidt in this ruzti : ddhindme “The Divine Book”. iwitdtndnw "The Book of Hyleg (i c apheta. the giver of life in u nativity)”, asrarndme “The Book of Secrets mazhar “The Manifestation", and <»l xosroundme “The Book of Xosrou" 3 4 . In this connection we must also mention rousanmndnt? “I he Book of Enlightenment ' hy Naser Xosrou (1003-1088) and sa'iddatndmr “The Book of Bliss” by another Naser Xosrou (d. 1352). A> golsan c rdz “The Garden of Mystery", an esteemed treatise on mysticism by sj - — — J y-*- Mahmud e Sabestari (d. ca. 1320). oik# jrjSf oi-kdseqin “The Lovers' Qualities" by J5U Helali (d. 1529). and finally two remarkable masnavidt by the most original of the Indo-Persian poets. Bide! (d. 1720); the philosophical o^o- telesm c fwirat “Talisman of Amazement ", and the travelogue ci/*- j^r fur c rnakrefat "Sinai of Knowledge". (Rypka 237-238 227-228. 188-189 186. 216. 186 fn. 254/246, 500-501/276, 518 407 ) § 1% Though primarily a mama vi -met re, this vazn is also common in other forms of poetry, qazaliat, qasd'cd. etc., and the dialect quatrains hy the eleventh-ccnlurv poet Baba Taher e SOryan are written in this rhythm (see the example in $ 194), Otherwise quatrains are generally written in vazn e roba 57 (see $§255-260), but also JUI Mohammad Eqbal (Sir Muhammad Iqbal. 1877-1938) uses hahr e hazaj e mosaddas c mahzuf for his quatrains. References: Gl 16. 27. 31. 50. 109. 117. 125, 135, I5i-I6i. 186. 239. 290, 314. 326; GII 3-7, 17. 29, 88-90. 96, 117, 133, 135. 143. 151, 241. 300y GUI 21. 27-50, 81-83. 91, 117. 135, 144-145. 194, 263. 267 his, 269. 286, 287-290, 349*; H 55, 99, 130. 137, 162. 217. 245. 251. 279. 282. 3 23. 33). 332. 386. 389, 419. 428. 431. 438, 447, 460. 463, 483; S 467 (Q -am, bus t). 496 (Mr -Cut tisi). 507 (Ml -d/r). 544 (T -drat, ntapcnddr), 545 (T -ast, nasavud), 558 (T -Hast). 565 (T tut hast). 581 (1 -uri naddrud). 587 (T -am nabusat !) , 588 (T -in nabasud). 598 (T -dnand, oduvttndan), 608 (T -dyad, nuga/iam), 618 (T -«r), 623 (T -its, raid). 634 (T andam), 644 (T -dnam, mard), 646 (T -Jim, na a:), 651 (T -drim), 655 (T -i ban), 659 (T -,uan), 676 (T -ardi. ntapors), 701 ( I -fir. go! ast). 715 (B -an ast, ce ray), 735 (B -a \ e to basad), 745 (B 'dyad, it .van). 752 (B -as. qidmat ). 752 (B -us, yek i). 754 (B -am. rqftq). 761 (B -ardn), 765 (B man a:). 767 [B -at. to bd in), 768 (B -di. garam). 769 (B -db i, ke dast). 788 (X -drd). 808 (X -an), 825 . / J 1 oyl/Je leJ/ ^ ^ ijj L*AP{J i ey falak d- lies tv mu ka- r I m asa » salt ' v ira n karde i moik v fra** rd he niarg e ia/r 1 vira n karde i "O heaven, go quietly! You have done a work (which is) not slight: By the death of the shah you have rendered desolate Irarfs land"\ Note how the rokn is marked off by the use of internal rhyme; malk v fra t! sd/r' vtra fl In the oldest period this vazn is not too commonly met with, but its popularity increases steadily throughout the classical period until m the post-classical period it has become the most common of all ouzdtt employed in qazatidt and qusd'vd. Roughly one sixth of all qasd ett are written in hahr v ramal v nwsamman v mahzuf. References: Gl 55, 114, 115, 122, 131, 207, 200, 242, 267, 323: Gil 26, 28, 76, 97, 194, 195, 220. 226. 228, 231, 248, 255, 262, 265, 271, 275, 297; Gill L 13. 15, 17 hh, 22, 23, 35, 48, 50, 68, 70 bis, 78 his , 98, 102. 106, 108, 112, 120, 122, 124, 125, 126. 127, 132, 5 Or v dn osar . 6 I.e. the beautiful locks of the beloved with which our heart is bound. 7 xuh and xuhi also mean “beautiful" and “beauty" respectively, and in this heit have to be understood in these senses as well. 128 THE RHYTHM 145. 146, 150. 151. 152, 161, 169, 170, 171. 183. 184, 189, 190, 192, 204, 207, 237. 243, 254, 280, 293, 315, 331; H 10, 12, 14, 31, 43, 62, 71. 77. 83, 92, 169. 199. 206, 212, 218, 240. 255. 265, 267. 276, 286, 309, 346. 349, 352, 390, 401, 402. 410, 433, 470; S 503 (Mr -/>/), 509 (Ml -ar). 537 (T -uz rd ), 541 (T -//>, cy nursalmandn). 552 (T -dr a mu da si), 557 {T -('/ asu duty), 561 (T -u \r m ha ), 563 (T -w st . kas), 564 (T * it iv dust), 566 [T -dr nisi), 571 (T -ah ddst ), 579 (T Hint), 591 (T -an goftt* antf), 606 (T -ar mfttmuf). 61 1 (T -dr. id ), 616 (T -ii\ Jiune). 622 (T -anas, tun), 633 (I -di zudam), 647 (T -int), 648 (T -ah ufkande tm ), 648 (T -due im ), 662 (T -an nn ikon), 666 (T -an ajkandc ih 673 (T -as if), 674 (T -d bar das ft), 687 (T -an mihari), 702 {T -iJv, vutjt), 708 (B -ir m). 710 (B -dr ru), 736 (B -di kysad), 750 (B -cm i yd mahds), 753 (B -dr c x H (s), 760 (B -an dsude in ? ), 762 (B - an has tan), 764 (B -in v man), 766 (B */ m vc to). Ill (B -udi kdski), 779 (B -dhat tlidamf). 788 (X -js rd), 789 (X -dr x"js rd), 792 {X -any nisi), 800 (X -due has). 801 (X -dr c did), 809 (X - d ve ro) % 821 U saldmun o absdl by Jami (d. 1492). and two short nuixnaviar, } jL nan o ha/vu "Bread and Sweetmeat" and j Ot tutu o panir "Bread and Cheese", by Seix Bahai (d. 1622). As late as fifty years ago there was composed in this vazn another mystical masnavi, which may claim a place in world literature, namely jdvidndmy "The Eternal Book" by the last great Indo-Persian poet, Mohammad liqhal. (Rypka 284/274. 287/277-278. 300/290. 732/381.) §203 In qasd’ad and qazatidt. too. the bahr <• ramal e musaddas v nuthzuf is quite common. References: Gl /. 25. 54. 103, 324, 327 bis; Git 18. 38. 40, 75. Hi-87, 96. 117. 12M31, 9 When talking about a masnavi the term bahr c ramal is sufficient to describe this vazn. since no other variety of the bahr e ramal is commonly employed in mas/uividt, but cl §§213-214. 10 Cf. footnote 3 above. 11 Quoted in §§100, 112. 166. and 20!. Moulavi is considered a Turk by the Turks and a Persian by (he Persians. He was probably a Persian by race and language, but he came to Turkey at a tender age and spent his life there, and Rumi, the name under which he is best known in Europe, means "the Turk” (< "(East) Roman”). The Turks call him Mevlana. The esteem in which the book is held is expressed in the following saying about its author: b'h if r man re gqyam vasf e d" * id - U janab HfS M peiqam- bar wlf da- rad ketdb "What shall J say to describe that exalted master? He is no prophet, yet he has a Book "! 130 THE RHYTHM 133. 149, 153* 163. 166* 182* 299; GUI 2L HJ, 123. 128, 146, 176 bis, IR7-IHH, 27R % 287, 297, 302 , 304-30$. 309*312. 312* 3)3; H 8* 96. 103. 197, 308, 363. 369* 478; S 460 (Q -dr. ha. i), 477 (Q -6/* t'_r), 506 (Ml -am). 511 (Ml -am). 540 (T -ah, tfdfef ami). 541 (T -ah, mdhruyd). 556 (T -asti x^ottar as!), 557 (T -f7 ast, pa). 567 (T -an is nist). 567 (T -am nisi), 592 (T -nty ami). 594 (T -ar mfzanutf). 595 (T -di mizanad), 596 {T -at mikonad), 596 (T -dni mikonud). 604 (T -a miravaii), 653 (T -an), 675 (T -ed ;)* 679 (T -ar \. jour). 681 (T -art\ mrv), 682 (T - ar L imr goz), 700 (T -t/\\ x"dham), 702 (T -a miravi). 725 (B -dr az dasi raft). 778 (B -dregi), 793 (X -anat). 802 (X -dm), 823 (Qq -d mfkonad). 823 (Qq - an mikanad). 925 (H -an e man). 926 (H -a w man). 927 (H -ar i, x w ps)i M 170*182, 264, 302-304, 424-432, 522* 810-832, 1095-1 107. 1208-1210* 1255-1260, 1309. 1320. 1326, 1346-1347* 1656-1676, 2005-2022* 2223-2231* 2380-2385* 2893-2926, §204 Other on zdn traditionally classified as varieties of bahr e ramal are described in §§209-214 and 248. §205 bahr e rajaz “the smarting metre” The basic unit of this metre is the robu mostafielon — and the only variety commonly met with is the standard version, bahr e rajaz e mo summon e salem "the sound eightfold smarting metre". It is practically always doubled ’ 2 . aknu" ke tan- sang I hezan giram nadd- az guse ye ha didamat talx i bogu ri met'/ e man cesm i be man ■ i i ™ ~ lot/ ar na a- tiq j hekes Vr mardom e •r nazzdre i t ■„ zdr i hokon 4 kar t hokon * cesmam gah i bari hokon 12 CT. §§ 186-187. §207 SECOND CIRCLE: RAJA / 131 “Now. since I have found (lit. seen) you alone, show me kindness (or) if not (then at least) tease me! Throw a stone (at me)! Say something harsh! Draw your sword! Do something! 1 take it that you have no fcclination towards me. Oh, light of mine eye (lit. pupil of my eye)! Some time, throw at least a glance at me from the corner of your eye". Ahli The name of the metre, “the smarting metre", goes very well with the content ol this example. §206 A famous qashie written in this not very common metre is the one by MoSe/zi cr sar'hci” ntonzel makon joz dar did • r li e ydr c man id \ek zamd'i zart konam bar rahS o at- fdi d demon & *■ 'O camel -driver, do not halt except in the land of my beloved one. so that 1 may once wail over the abode and the ruins and the traces!" References: GI 236: Gil 163. 167. 200. 230: GUI I. 25, 30. 34. 67. 73. 93. 138, 201; H 191, 344; S 536 (T -dh rd ). 536 (T -d: ru). 537 (T -dm rd. erm ah), 537 (T -dm ru . bar xfz). 582 (T -d miharatl). 605 (T -doom mhuvaif). 619 (T -ax), 628 (T -am), 649 (T md nt: ham bud nixtim), 740 (B -or bar kanad), 756 (B -an miharam). 771 ( B -art, aver); M 1-35. 261, 321, 519. 523-542. 1015-1019. 1172. 1215-1216. 1333-1334. 1370 - 1391 . 1777. 1786-1812. 2130 - 2140. 2275-2282. 2427-2453. §207 Other ouzdn traditionally reckoned as varieties of bahr e rajaz are given in §§215 and 250-251. 13 Cf. §§92 and 93. XIX FAST METRES (THE THIRD C IRCLE) §208 The metres of this circle all consist of two short syllables alternating with two long syllables, and whereas the metres of the second circle may well have been adopted from the Arabs, the metres of this circle are certainly of native origin 1 , and therefore they are not recognised by the traditional prosodists as independent metres, but classified as varieties of the hohur "metres" of the second circle. The first of the two shorts (never the second) is thought to have been originally long. In other words, in order to arrive at the traditional classification of the metres in this circle one should always begin by altering the first short to a long. §209 bahr e ramal e max bun “the hemmed running metre" The basic unit of this metre is fascia ton w--. It is traditionally not recognised as a separate bahr, but said to be a maxbun "hemmed" version of bahr e ramal, fascia ton - w- - having become fascia ton ^ ^ — . The term maxbun. taken from the jargon of the tailors, means that the first syllable of each rokn has been changed from a long to a short, Hence the metre described in §172 We learn from the scholia in the Alexandrian grammarian Hephaestion's manual of Greek metres that the latitats measure (two long syllables alternating with two short ones) was employed in Middle Persian poetry: 6 kcu nepaiicd^ itovtKo; fiAv, ou oi Rove? uOt(E> CTCCXpqvTO' TTcpcmedq 6 £. 6uj to ru^ iatopiu^ xuz IlnpoiKOs foimo pftpep ycyptitpOai. and the Persian one (it is also called); Ionian (it is called) because the lomans employed it, and Persian because the Persian epics were written in this metre". See Hephsestionis Alcxandrini Emhivuium dv mans et poemate (ed. J.C. Pauwf Trajecti ad Rhenum 1726, p, 82, Cf T Otakar Klima: Middle Persian literature in Jan Rypka; History of Iranian Literature , Dordrecht l%8, p. 53, (p, 54 in the German version). Cf. Additional Notes, Cf, the rhythm of the archaic dim ye kuhi quoted in §258, §210 THIRD CIRCLE: RAMAL E MAX BUN 133 \s \j uwO is said to be a maxhun version of hahr e motaitarek (§190), that is fdSehn v has hecome ^ , Accordingly the vazn of Seix Bahai’s masnttvi is hahr e motadarek c mosamman e maxhun. §210 By repeating faSefaton - four times in each mesraS we arrive at hahr e ramal e mosamman e maxhun “the hemmed eightfold running metre". zahed ar ah- joz be duzax ei x"os a n del Z nasavad si- l e behest as- co man i zol - « ke ze torkd- f 'te ve ran- m r *' r t •’ xoddva * m ' hovad vd- M n e pariceh- g i yo dso f- maferestam r e co it i w i re co yaqma te re bit i * v - ^ -t V u — “ \J — \J “If the bigot 2 is (destined to become) an inhabitant of Paradise, O God, do not send me (there)! (Do not send me anywhere) except to Hell! (For) someone like me the torment (of Hell) will be a friend like him (or her whom I love). How happy that heart which does not become infatuated with a colour 3 4 and enamoured with a smell 3 , as Yaqma (does) with fairy-faced Turks ’. Vacjma ye Jandaqi (d. 1859) It should be remembered that may be substituted for ^ in the beginning of a mesra *• C har samarqan- d } agar hag - name ye ah- / e xvrasd* of Khorassan" by Anvari (d. 1189?) zun ey ha~ d e stahar he har e sot* fd u har "If you pass bv Samarqand, O morning breeze, carry the letter of the people of Xorasan to the sultan". The elegy by Xaqani (d. ! 199) over hjs wife and only son beginning; a* •* •# deinavdz e man e hima- r’ .fomai- d* hattw 5 "Comforters of afflicted me are you all". 5212 THIRD CIRCLE : RAM At. F MAX BUS 135 ! lie ijitzul by Hafez beginning: f sohh dam mar - ftdz* kam kon q c caman ha ke {kit i fl ha * got e ruutxch t f* ha s t ( if 1 * v 'te go ft to si'koft "At duwn the field-bird {x nightingale) said to the new-groun (lit newly-risen) rose; "Be less coy for in this garden many have blossomed like you' References: Gl 63, 99, 101, 107 his. 206, 244, 268, 279. 283, 304; Gil 20, 23, 52, 53, 69, 137. 152, 161. 164. 168, 172, 173. 175 his, 181, 196, 199, 207. 208, 209, 211. 213. 214. 223, 225, 233, 236, 238, 239, 241, 242, 243. 245. 246, 247, 253, 254. 258. 261, 296; Gill 3, 14, 33, 36. 37, 40, 41. 45. 46, 59. 72. 95, 96, 100, 105, 112, 113, 115, 119, 136, 152. 166, 177, 178, 186, 191, 193, 194, 197. 199. 211, 216, 222, 226, 231. 238, 279, 281 his. 303, 343: H 9, 17-21. 24, 26, 36, 48. 49, 51, 52, 57. 67, 68, 70, 73-75. 80, 81. 85, 105, 108, HI, 112, 123-125. 128, 134, 136, 140-142, 144, 150, 152, 157-159. 164, 166. 173. 174, 176, 178, 181-185, 189, 190, 193, 202-205, 207-2) 1. 213. 222, 223, 227, 228, 236, 248-250, 252, 257, 264. 268, 271, 277, 281. 287. 289. 293. 301. 310, 311, 314, 316, 317, 319, 324, 326, 328. 335, 336, 340, 341, 345, 347, 348. 355. 357, 359-361, 366-368. 373. 376-378, 380, 385, 3K7. 391. 404, 407, 420. 422, 423, 448-450, 455, 456. 458, 467. 472. 480, 481. 484. 485, 488, 490; S 443 (Q -d bar .xdst). 445 (Q -uni nisi, ayyohd n-nus ), 450 (Q -ar ha: donut), 452 (0 nitty 454 (Q - dr , hanutad /), 538 (T -an ra\ , 543 (T -at at). 547 (T -a r dnja .r/h 548 (T -a mat bar ,\ast), 548 (T -ah ast ), 552 (T -as i oftddast), 555 (T -ar ast , har). 563 (T -ant e i hist). 564 (T -am a: ft .v/), 565 (T ~ar f hast). 566 (T -er nisi). 568 (T -dmam ntst). 568 (T -an nisi). 569 (T -dr e to nisi), 570 (T -dtt e to nisi ), 573 (T -dr heraft), 581 (T -an j ddrud ), 583 (T an tnigozaratf), 584 (T - d sod), 589 (T -ar misod), 591 ( T -an nidnatf), 593 (T -aramt). 593 {T -draml). 593 (T -d/as daramt). 600 ( V -tui, saraf). 600 (T -d hud )* 603 (T -et heraviuf), 603 ( Y -d naravad), 604 (T -tin naruvad), 606 (T -id), 608 (T -d: dyatf). 609 (T -dq dyad). 614 {T -dr, he faluk). 617 (T -ir, /nd), 622 (T -dzas), 625 (T -is), 631 (T -as tarn, man). 632 (T •nftanu man), 635 (T -an hudam), 638 (T -a minegaram), 639 (T - d:um , ar). 639 (T -dzam, nazar). 651 (T -an gardidim), 653 (T -d narasim), 661 (T -an didan), 665 (T -dr/ heh). 667 (T -dr). 668 (T -dr di). 684 (T -dr L xahar ). 689 (1 -anas hast). 693 (T -fa)m\ man). 698 (T ap koni ), 708 (B -an rd . et ke)* 709 (B -di rd). 714 (B -dktar ast) 717 {B •in ast , gar), 722 (B -di hast , har ee). 724 (B -hi i st), 726 (B -d hegerefi), 727 (R -di al), 730 (B - in mjgozaratf), 733 (B -as i har xizad), 733 (B -d tuna suit), 734 (B -dnam hasad), 737 (B -an hdz annul), 740 (B -an to cand). 74 MB ■■■ - of nakonad), 741 (B - amand ). 743 (B -d dyatui), 744 (B -d ye to hovatf), 745 (B -u heravad), 747 (B -tin midyad), 748 (B -ak i midyatf), 748 (B -dr v degarh 753 (B -is), 754 (B -d ye to am), 755 (B -ddamh 758 (B nmam). 760 (B dm). 762 {B -ardan). 767 Adib Saber ft del fun- gi rat! rah su-ve ki*x- * v ar v digar girad -J v> He must lear hi.\ hear! away from existence and take the road of another land dsiati du- st* dar a » du- num c dust zdqak / zvs- L. salamo 5 aleik has to be scanned ^ . Apparently is was read saldmwdh-ik with a falling diphthong in the third syllable as in French moi. XX COMPOUND METRES (THE FOURTH CIRCLE) §226 The ouzan treated so far are each made up by repeating a basic unit or rokn a number of times with or without minor modifications of the basic unit. The ouzan in this circle are each formed by two alternating basic units, in most cases mafaSelon o- alternating with faSelaton ^ ^ — . By the traditional prosodists these metres are more or less artificially considered to be varieties — maxhun, maivi 1 etc, of certain compound Arabic metres, §227 hahr e mojtass "the chopped metre" In Arabic prosody the basic arkan of hahr e mojmss are mostafielon — w- and fa^eiaton - ^ . By making both of these maxhun , i.c, by shortening the first syllable of each, we arrive at the basic units of the Persian metre, of which the standard variety is hahr c mojtass e mosamman e maxhun “the hemmed eightfold chopped metre". 1 ft ? h I Cf §§209 and 215. $228 KOURTH CIRCLE: MOJTASS 145 hezar' da- v hezar' san- *■ sezad 'agar ke mqd e Cani- ne fesat' don- g'' be bdlam naforusam r' ze kaf dd- d o ramesa" zadi ro man w qam e lord dam d qam e “ \j — nasodam man naparidam be do V dlam * to xaridam w w ‘They scattered a thousand grains, but (lit. and) to them I did not become tame. You hurled (lit. struck) a thousand stones at my wing, still (lit. and) I did not fly away. It is meet, if 1 do not sell the anxiety (for the sake) of you (even) for the two worlds, (considering) that I paid the cash (consisting) of (my) life and bought the anxiety (consisting) of you". Sabahi (d. 1803) This vazn is rare. References: Gil! 164, 179, 217; H 234, 322, 469; S 442 (Q -in m), 510 (Ml -of), 511 (Ml -afi), 543 (T -dtlaih 576 (T -an aO* 608 (T -ur dy(uf) m 633 (T Htnlum), 635 (T -Warn), 640 (T -a ye to bus am), 641 (T -mum)* 671 «, 250. 251, 252 hix, 264, 275. 277, 284, 286. 300, 313; H 2, 4. 16. 22. 23. 25, 28, 32. 34. 35. 37. 41. 44. 45, 47. 50. 53. 54. 58, 61, 64. 66. 76, 79. 88. 93, 97, 98, 101, 106, 113, 114. 116, M9, 122. 129. 131, 132, 135. 143, 145. 147. 155, 156, 160. 167. 168. 175. 177. 179. 187. 188. 195. 201. 215, 219, 221, 224. 230, 235. 237-239. 241, 242. 244, 247, 256. 258, 259. 261. 263. 266. 269. 273. 274. 280. 283, 290. 292, 295. 297-300, 303, 305. 306. 315. 330, 333. 337, 339, 342. 350, 358, 379, 388, 393, 397. 399, 403. 405 . 416. 421. 427, 430, 441-443, 445, 446. 452, 457, 461, 471, 477. 491; S 444 .?), 447 (0 -ait). 450 (Q -an monad). 452 (Q -bzand), 456 (Q -dr. he hie). 458 (Q -or. kojd), 462 (Q -or, .Void), 462 (0 -dr, sab i), 464 (Q -el, liar), 466 (Q -d/). 469 (Q -am, xoddy rd), 473 (0 -do, tamam), 473 (Q -an, . iakar ). 476 (0 -do. ford). 478 (Q -in, taharaka Hah), 482 (Q -ah). 483 (Q -dr). 484 (Q -di). 487 (Q -ddi). 487 (Q -art. he: an), 489 (Q -art, gar). 498 (Mr -axr). 500 (Mr -ml). 501 (Mr -antiax), 501 (Mr - arax ). 506 (Ml -dhi), 534 (T -did. agar). 535 (T -d rd. x ah), 539 (T -di rd). 545 (T -ust. rood n). 553 (T -ami as!). 557 (T -ung axt). 560 (T -do an). 562 (T -asr, hold), 562 (T -ust. safar), 571 (T - angi oft/). 575 (T -at tut. t o). 578 (T -dr dar nuntigonjatf). 580 (T - drad , ke). 583 (T -dztttf). 590 (T -amid, kas f). 592 (T -a stand). 594 (T -izarni), 595 (T -ixand), 596 (T -us nakonad), 599 (T - ud , agar), 607 (T -dyad, he hash). 610 (T -ir mjdyad), 622 (T -anus, ro/id), 625 (T -dq), 627 (T -«/, moo), 628 (T -ul. neiasle), 629 (T -dm, to), 629 (T -am, hekdyat), 630 (T -dm. mara), 631 (T -aslant, he xdk). 636 (T -arum). 637 (T -dram), 641 (T -dr k giant), 646 (T -inam, man), 661 (T - ardart ). 677 (1 - andi . ce), 684 (T -ari. do). 685 (T - dri , man), 688 (T -dr/). 690 (T ftSi), 691 (T -ami). 699 (T -ini. sab), 709 (B -u rd), 711 (B -axt), 711 (B -do andaxr), 711 {B -ari amuxt), 712 (B -d.r/). 715 (B -do axt. hezdr), 716 (B -dr e man axt), 718 (B -ir or dust). 720 (B -d ei dust), 720 (B -do ei dust), 722 (B -d / hast. hid). 723 (B -d / si, zeh f). 723 (B -d j st, mara), 724 (B -angi nisi), 725 (B -ar migaxt, xidf), 725 (B -or migast, del f). 728 (B -d gonjad), 728 (B -dr bar gardad). 729 (B -or ddrad), 730 (B -dr e man ddrad ), 731 (B -ar girad ), 732 (B -dr bar namtgirad), 734 (B -do ce re sad), 736 (B -an bekosad). 6 Still it accounts only for slightly more than one twelfth of his qazalidt. The rhythms used by Moulavi show great variety, and he often employs rare ouzdn. 148 THE RHYTHM 740 (B dzand), 742 {B -dm konatuf)* 742 (B -avand), 743 (B -ud, ford), 746 (B -dyad, ferdq), 746 (B -dyad, marou), 747 (B -d namidyad), 749 (B -di bdz ), 750 (B -amt, x»oi) 753 (B -ar v x "js\, 756 (B - andam K 763 (B -idan), 764 (B - d I ast in)* 768 (B -di , darky), 77 f (B •uiidi), 774 (B -an art). 775 (B -an dart). 775 (B -uri), 775 (B -dr/), 776 (B -as /, hami zanam), 782 (B -an hvgarddni), 785 (B -ar cc ntfX^ahf), 794 (X -d ce qam ddrad), 795 (X -dr e md hdsad), 797 (X -dr mkidncuf), 797 (X -ar girrndU 799 (X -dr dyad), 800 (X - ur ), 801 seiseiat oz-zahab "The Gold Chain" composed in 1485 by Jami, and ^erfan “Gnosis", Bidcl’s principal work, written in 1712. (Rypka 236-237/225-226. 254. 246-247. 286/277. 517-519/407-408.) Of a quite different sort are _»> haft peikar "Seven Portraits", the last of Nezami s masnaviat. composed in 1197, in which he describes the exploits of f'jr Bahram e Gur, and its celebrated imitation hast behest "Light Paradises" by Amir Xosrov e Dchlavi (d. 1325). Mention must also he made of ^ >j>) *’Li sdh a Jarvis "The Prince and the Pauper” by Helali (d. 1529). (Rypka 211-212/204. 258/250, 285-286 & 500-501/276.) References: Cl 7 his. 8 bis . 15. 16. 51, 52, 64. 118. 128. 172. 221. 222. 224. 257-261, 268. 273. 274, 298, 308, 313; Gil I2-i3, 42. 78, 115. 149. IH6 bis, 218, 219 bis, 262, 276. 295; GUI; 43. 59, 148 bis, 154, 244. 258; H 13. 42, 56. 262. 270, 302, 381, 453; S 463 (Q -al, sokr). 472 (Q -an, barg), 485 (Q -abi). 543 (T -axi), 544 (T -drat, bandfvar), 569 (T -idan nisi 1 * ). 570 (T -ir e to /list}. 571 (T -an dast). 584 (T -or bdsad). 595 (T -dr konad). 617 (T -dr, ci). 617 (T -dr, motaqaHeb). 619 (T -£r, sdqi id). 621 (T -dras. har ke ndzok). 621 . 726 (B -n/mi. o/i/r/w). 729 (B -art/). 750 (B -abas). 751 (B -«««>•. Aar ke hast), 772 (B -«r f, didam), 779 (B -a/j /, ccWe). 780 (B -an /. lor), 785 ( B -an‘). 790 (X - in the above example, the term y jhZ* maqtu $ “cut” describing the “change” from — ^ - to . §234 In this not very common va:n is written the famous poem by Rudaki (d. 940) describing the production of wine beginning % •« m H * nttidar e met # ra bekard' hdyad qor- ban haccf vc u ra geref t o kart t* he zen- dan ”■ — — “ — 23 Cf. §98 end. FOURTH CIRCLE: MOZ.iRES 155 §238 This rhythm is then made axrab — the first rokn is changed to - - « (Cf. §220) — and makfuf— the final long syllables are shortened in all arkan :4 (cf. §220) — and mahzuf Thus we arrive at the rhythm of our example, which is therefore classified as bahr c mozareS e axrab e makfuf c mahzuf "the ear-pierced restrained apocopated eightfold similar metre”, but of course it is generally referred to simply as bahr e mozdre V, §238 This is one of the most common ouzdn, ,md the one which { personally like best. 01 the many well-known poems in this vain* we shall only mention two: The beautiful qazat by Moulavl (d, 127?) beginning: %-wr 1 rox ke ha q a Uth ke qatui e gotextamh m areiu si faravmuh m arezu st “ U w V ” V V “ V ' — w 1 ■' "Show your face, for I desire orchards and gardens! Open your tips, for I desire abundant sugar! The very popular tarkibbami 2S on the martyrdom ol Le-— *- Hosein by jLitf Mohlasam e Kasant (d. 1587) beginning: hdz i» ce x ures a si’ ke r tar xalq e Saturn asi bdz /" ce noube no ce kuza mi ce mat am asi “What commotion is this there is once more among the people of the world? Once more, what lamentation and what wailing and what mourning is this"? (Rypka 298/287-288.) References: G! 5. 34. 41. 42. 49. 58. 62 bis, 63. 64. 65. 104. 125. 141, 175, 188. 207, 215. 249, 261, 273, 278, 280, 291, 298. 300; GI I 28 , 41. 42, 43 , 58 . 75 , 77 . 78 . 94, 99, 103, 112. 131. 134. 144, 174, 176, 182. 189, 193, 197. 220. 221, 256. 266. 284. 298; GUI 4. 10, 16, 22, 35, 57. 59, 71, 74 bis , 79. 80, 87, 88, 90 his. 94, 99. 104, 105. 106, 113. 114, 121, 135. 139, 147 bis. 162. 163. 198. 203, 206. 210, 213. 215. 235. 240, 242, 276. 283. 285, 291, 294. 298, 314, 317-326, 334. 337; H 7, II. 30. 33. 39. 59. 60. 63. 65. 72. 78. 84, 86, 87. 90. 91. 100. 102. 133. 138. 139, 180. 186. 196. 198, 200. 214. 220, 225. 226. 229. 243, 246. 253. 254. 260. 275, 285. 291. 312, 313. 320, 321. 329, 338, 343, 351. 353. 362. 364. 365, 372, 375. 394- 396, 398, 400, 406, 408. 409, 413. 415. 424. 429, 437, 439. 451. 459. 465, 479. 482, 486, 487; 24 This, of course, does not apply to the last rokn where the (Inal syllable has disappeared completely. In the first rokn the last syllable is already short since it is a.xrab. 25 Cf. § 144. V-J 156 THE RHYTHM $238 S 438 ( 0 -«)* 448 (Q -dr leant). 470 (Q -tin, in ), 48 [ (Q -f re to). 489 (Q -art). 492 (Q -el /). 499 (Mr -an heraft), 538 (T - dh rj), 539 (T -ast e mo). 542 (T -/6 2t \ rafti). 551 (T -an v to .vf). 554 (T -ar ast, az h 555 (T -ar ast. in). 556 (T - dh x^psiat ast). 558 (T -am ast), 559 O -an ast), 563 (T -an e dust). 564 (T -dmn ftrezti .vr). 582 (T -dr bogzarad), 597 (T -anand), 601 (T -ft bovacf), 602 (T -ft r avail, hesytir ), 602 (T -d raxaiL fib f), 606 (T -ak i sauuf), 610 (T -dr hengarhf), 614 (T -d ye ydr ), 618 (T -iz, peivamf). 620 (T aha las). 620 (T -uv), 627 (T -til, hide!). 634 (T -ar sodum). 643 (T -anam, gar). 651 (T -arim), 652 (T -u konim, bar \rz hi he Sahd). 652 (T -ft konim, bar xiz id htrit/ ), 661 (T -ar soxan h 665 (T -dh), 674 (T -drat tf). 678 (T -mi. t i burp). 681 (T -ar L kas). 683 (T -an. bar nmbatam ), 693 (T -ant. us tide). 693 (T -ant\ ptikize), 698 (T - ar kiwi). 707 (B -fib rd), 714 (B -an e to st), 714 (B -wr x u p\tar ast K 717 (B -dr e u sty 717 (B -dm e ft st). 718 (B -dy e ft si). 719 (B -a! e dust), 719 (B -dm e dust). 720 (B -tin e dust). 721 { B -ft ye dust, stidi), 724 (B -ib njsf), 726 (B -ar bar gereft), 728 (B -am tiff ad), 730 (B -a ke hard h 737 (B -dm sod). 739 (B -hie and). 744 (B -ar ho wit). 765 (B -and e (0 766 (B -sty 0, 773 (B -art. rajti ), 782 (B -fir mikoni), 783 (B -tin begin), 791 (X -a! e dust)), 793 (X -ar fetatf), 798 (X -dr mtkonad). 799 (X -tin saratl), 799 (X -arum ei dust dost gir), 805 (X -ar mikonam K 810 (X -id i nay aft e), 815 (X -fc mfkoni). 819 (Qq -a ye to st ), 820 (Qq -dial ast), 820 (Qq -am asi). 821 (Qq - til e dust K 826 (Oq -a tv a ,1 /y), 826 (Qq -dp), 831 (Qq -ine f), 919 (H -ddan ast), 923 (H -as, Somrat). 924 (H -as. hum/ens). 924 (H -tip) '. M 197-203, 308-310, 441-459, 861-880, 1116-1122, 1 198-1 199, 1268, 1298-1299, 1348. 1704-1713. 2044-2054, 2233-2241. 2400-2401. 2972-3006. §239 By reducing the above vazn to eleven syllables we arrive at the very rare masnavi- -variety: bahr e mozdreH e mosaddas e ax rah e makfuf e sdfem 27 "the ear-pierced restrained sound sixfold similar metre". ei d" ke v andar ne- raft d“ ke bud a" ke — “ \j ( w cjamgent vo ha“ sere.sk '' raft o dmad bud' xire * ^ V “ w “ v sezdvdri m hami bdri v dnk dmad ce qam ddrf yj — — “ — — third mesial) 26 The rhyming element in this pazal is a good specimen of entitle* Cf § 145, 7 The order of the terms is important; the first, second and third rokn arc axrab, makfuf and salem respectively. §240 FOURTH CIRCLE: MOZaRES 157 "O you (lit. he) who are (both) sad and worthy and weep tears in secret! What has gone has gone and what has arrived has arrived, What is past is past. Why do you regret groundlessly"? Rudakl (d. 940) According to the traditional terminology the second he if should be described as huhr e mozdreS e mosiuhlas e ax rah e sdlem e maxannaq ou axrah e makfuf e sdlem* the term jJe moxutwaq '’suflocateiT' indicating (he change of v to in any rokn except the first where a similarly modified rokn would be termed ax raw "pierced through the nose'\ Ct §254. Reference: G! 13, §240 Closely related to the last two ouzan is the doubled metre of the following example « ** * h L If ifj 9 */ ** l* If/ ^ J f' M 4 ** * jouq i qa- bud 6 to- most i mo- m gar hice- kindardnim jud e mdrd jarraddrutn ma" nabdsad bar ma qa- hak az Va- bar faqr ' az hie ’ lam nabdsad dam nabdsad del nehdde * qam nabdsad “ “ w ” \j ' | w w ~ “We are a group of vagabonds and have no responsibility (lit. on us there is no pen). Our being and existence has no fear of poverty 28 . We are a handful of free (men), our hearts set on poverty. If we possess nothing, we worry about nothing". SObeid e Zakani (d. 1371) This vazn, which is not too common, must be termed bahr e mozdre 9 e axrah va Salem. References: Gl 59. 176: Gil 95, 136. 223, 268: GUI 9. 134. 256: H 5, 94. 126, 154. 171, 231. 233. 307. 392, 425. 426, 432. 434. 435. 444. 462, 464. 489; S 576 (T -mat, x" V s\. 579 (T -drad, dtddr), 585 (T -ar nabdsad hd)* 588 { Y -in nabdsad, gar), 656 gzdr h 658 i I -dndn), 667 (T -ane), 675 (T -ddh ei). 683 (T -art, cun), 685 (T -dr L * iomr i), 688 (T -irt), 692 (T -ami, sdhehnazar), 695 (T -ant. zouq i), 707 {B -a rd , moSiaqi), 732 IB -tin zad ), 747 (B +um bar dyad), 778 fB d! t), 784 (B -ah f), 795 (X -ar hehdsad), 829 (Qq -dr 28 Sudani also means "privation, loss* annihilation* 29 Note the use of va, cf. §223* and these meanings are intended, too. 158 THE RHYTHM $240 daraml. M 185-1%. 305-307. 436-440.839-860. 1111-11 15 1211. 1261-1267. 1310. 1685-1703. 2027-2043, 2386-2399 . 2931-2971. 3194-3195, §241 bahr e qarih ‘‘the close metre” At this place mention should be made of a rhythm which, though not fitting directly into Circle Four, bears a certain likeness to the rhythms of this circle, being constituted of the same basic elements, namely - and \j ~ w yfk muy* bedozdidam az do zolfat cu " zolf' zadi ei sanam be sane fund's' be saxti hand kesidam * mm cu" mur ' ke garuiom kesad be xdne bd muy' be xdne sodani pedar goft mansur ' koddm ast a: i" do gdne “I stole one hair from your two tresses, when you were combing them, darling. I pulled it along with such difficulty, like an ant, who pulls (a grain of) wheat to (its) home. I went home with the hair. Father said: "Which one of these two is Mansur'?" rUt* J Mansur e Manteqi (d. ca. 985) This is traditionally analysed as maj'Sulo “ “ c/ majd Silo fa Seldton “ U " “ 30 No 1 1 12 has many Arabic isolations which Moulavl has allowed to lollow the Arab rules of permitted variations. §244 FOURTH CIRCLE: QAR/B 159 and said to be derived from the Arabic hohr e (forth , of which the standard mesrd $ is i/W c yy mafa ilon mo fa Won fa ( ieldton v — yj — — v Hence the above example is classified as bahr e (forth e mosaddas e uxrah e makfuf e sdlem “the ear-pierced restrained sound sixfold metre". Cf. §239 footnote 27. his vazn is rare. References: Gl 43, 48. §242 Another variety of bahr e qarih was seen in §158. That rhythm can be described as bahr e qarih e mosaddas e satem os-sa Jr * 1 va ax rah oh ebteiia va ax rah ou makfuf oh hasavem 31 va maysur * 2 oz> and the standard form of halir v jutliil as 4.5.04 08 ■ - J \ The method is quite ingenious. The only thing which might be objected to is that there is no break in these rhythms. §246 bahr e moqtazab “the extempore metre" Before proceeding to Circle Five mention must also be made of the very rare bahr e moqtazab. The Arabic standard form is w — ^ The Persian standard form is mafvi - w v-, Hafez has a qazal in this \azn, in which - has been substituted for w v throughout. Reference: H 473. 35 Elwell-Sution pp. 88. 104 and 107. Cf. §165 above, XXI COMPOUND METRES TRADITIONALLY CONSIDERED TO BE VARIETIES OF SIMPLE METRES (THE FIFTH CIRCLE) §247 Just like the ouztm of Circle III, the ouzem of this circle, too, are traditionally considered varieties of Circle II. Since in this circle compound metres are classified as varieties of simple metres, the traditional terminology becomes very cumbersome. As mentioned above (§ 162), Circle V is merely Circle IV read counter-clockwise. §248 hahr e ramal e maskul “the tied running metre” tan e ado - na hami" le- agar ada- ce miart e mi sari f as- has e zibd «• mi be cesm as- » H naqs e diva - /•’ be jan e Mamiyai s'-' nesdn e ddamiyat t o dahdn o gus o bini r a midn e ddamiyat ■ m r — " vy V “Noble is the human body by its soul of humanity. (It is) not this beautiful clothing (that) is the sign of humanity. If being a man is by (i.e, if one is a man by virtue of) eye and mouth and car and nose, what (difference 162 THE RHYTHM would there then be) between a picture on the wall and between humanity 1 ? SaSdi A rokn in which both the first and the last syllabic have been shortened is termed maskul "tied". Hence the above vazn is classified as bahr e ramal e mosamman e maskul va 2 sdlem “the tied and sound eightfold running metre", normally referred to simply as bahr e ramal e maskul. §249 It should be noticed that in this 10 ^ is not substituted for ^ in the first rokn, and likewise the substitution of for is extremely rare. Cf §152. This vazn is generally doubled* thus the example given in S 154. It is not very common. References: G! 103; GII 141, 160, 257: GIU 139, 204* 228; H 6. 1 17, 468; S 542 (T dbut), 577 {T -iyut). 584 (T -dz hdsud), 585 (T -d mtbusud), 620 (T Hindus). 632 (T -oftanu co). 637 (T *ah durum), 638 (T drum, he \odd), 656 (T -m/an), 670 (T -at, xabarut), 686 (T -dh dart. 687 {T -ah dun , to), 690 4 y Ji A **j {* sine ye ban - de jd ye to md home dar havd ye to * / ta rox e x"od nornude i i * at as e man fozude i o 1 * - w dide be zi - * ■ R to be havd « jd" ze tanam mehrjazd — \J sj ” r e pd ye to ye kis 'fi robude i ■ re kis'tt \j — * 1 It is a popular joke amongst the Persians to mispronounce the first beit and make a pause after na. The meaning is thereby changed to; "Noble is the human body by its soul of humanity. Oh no, it is not! It is the beautiful clothing, that is the sign of humanity”. 2 Cf §22.1. $252 FIFTH CIRCLE: \RAJAZ] 163 “The breast of my humble self (ought to be) thy seat. (My) eye (I want to place) under thy feet. We are all longing for thee. Whom art thou longing for? As soon as thou shewedst thy face, thou snatchedst away my soul from my body. Thou hast increased my fire. Whose lover 3 4 art thou"? Amir Xosrov e Dehlavi T his vazn is termed hahr e rajaz e mosamman e matvi va maxbun “the folded and hemmed eightfold smarting metre", since the arkan can alternately be considered as matvi and maxbun varieties of mostafSehn §251 This vazn, which is not very common, is often doubled as in the above example, where the doubling has been marked ofT by internal rhyme. References: 091 140, 183, 274; GUI 51. 69. 85: H 192. 296. 382, 411, 414; S 577 (T •dial), 620 (T -dmataS). 642 (T -amm, t?sm), 657 (T -iatt). 664 (T -w r< j i/>. 666 (T -axle /), 680 (T -ari, ddnamai), 681 (T - ari , rur), 688 (T -x i), 699 (T -a: mikoni)*, 736 (B -a! soil), 757 (B -clam, bar), 757 (B -e/atn, id), 763 (B -dl e man), 773 (B -ar j. gar), 780 (B -( ajni , kas), 784 (B -A/), 825 (Qq -dnamai): M 44-52. 322-323. 548-561, (020-1021, 1205-1206. 1220, 1301- 1302. 1336, 1402-1411, 182M843. 2147-2159, 2263. 2285-2289. 2464-2497, 3179. izf ■J*’ dl)s > cr u ' guy and' pestd " be sabhd ba- bidar ’ mard co za- dahan geref- r e gah Wa- nesast a xof- d ' mddar tan dmiixt # re ye man tan dmiixt » 3 Untranslatable pun: mehrfaza also means “mcreaser of fire‘\ 4 The last be it in this qazal furnishes an illustrative example of the utility of a knowledge of prosody: it would be difficult to understand it without knowing the vazn. 164 THE RHYTHM §252 "They say that when my mother bore me, she taught me to take the teat in my mouth. During the nights she sat awake at my cradle and taught me to sleep ". I jjy* \ fraj Mirza (d. 1924) This vazn is classified bahr e hazaj e mosaddas e ax rah e maqbuz e mahzuf* "the ear-pierced restricted apocopated eightfold shaking metre", the term maqbuz "restricted" indicating a rokn in which the third syllable has been shortened. §253 In spile of the fact that this vazn Only comprises ten syllables, a large number of masnaviat has been composed in it, the most famous of which is j JJ MU ro majnm b composed in 1188 by Nezaml. Of course JamI too has his /<■/// m mqjnw*. composed in 1484, and so has Amir Xosrov e Dehlavi (d. 1325) who preferred to call his version majnitn o MU 1 . but the best imitation is ihe admirable MU yo nuijnun by Maktabi. composed in 1489-90, (Rypka 210-213/202-205. 287/277. 258/250.) Mention must also be made of the mystical masnaviat* £*ajYI r meshah otarvah “The Lamp of the Spirits" by jj-y Ouhad od-Din (d. 1298) and ha/name “The Book of Ecstasy", better known as uL5>> j g«i o fimad), 613 (T * dr * iart), 615 (T -ur. parvdne), 615 (T -j>, an), 624 (T -w.v« raft! ), 626 (T - ang h 634 (T -udam, man ). 636 (T -ari na didam), 637 (T -ar datum). 638 (T dram , j gar), 642 (T -atom), 654 (T -an, ei) 655 (T -an, har xiz), 662 (T -a kon ), 666 (T -fide), 694 (T huu\ ei) 727 (B dinar, gar), 759 (B -anam. 5 Cf, §239 footnote 27. 6 Or hula no majnun. 1 Or mu j nun o leild. 8 I.c. Persian Iraq and Arabian Iraq, 9 Sec § 144, $254 FIFTH CIRCLE; [HAZAJ] 165 g or), 783 (B 790 (X -ast). 791 ' It;) gffjf Jr"' eJ W JvVj' J'Z b/' ,1 ;) (** » (f 'ff* mqjmu t ie ye kou- tv’ 6 ra be qd- nun e sabaq — — ^ w — kardinr' tasaf- — — O “ haqqa ke nax'd"- ' ^ v ■■ joz zdt e haq o v r W — fob varaqan " ^ “ dim o nadi~ “ w v “ sefat e zd- — V babda varaq — ™ dim 1 dar u ■ — “ r/ve ye haq — — V/ v w ^ — V* vy w “ “Like a lesson-book, the compendium oi the Universe We turned over, leaf after leaf: In truth we read and saw therein naught Save the Essence of God, and the Essential Attributes of God” \ 6 Cf. §88. 7 I do not know the author of this rohdSL h is quoted in E.G. Browne: A Year amongst fite Persians, 3rd cd, 1950, p. 149. The translation is Browne s. $257 THE ROBASI MtTRE 169 gar rtt ye zamr he jomle d- had' koni “ “ kj “ heh z d" nahovad \j ~ “ ke xdter i — W “ sad' koni “ " y u — gar ha tide koni W — w ” he lot/' d~ “ ^ “ zdd i rd ■ — w v — heh tar ke hezd- M “ \j — r' hande a- ■ zdd' koni u y M C )j s' ^ iS 1 ** dor Helm o honor mason co man sdheb e fan w w W “ v td nazd e Hazi- za" nasavi a "dr ' co man \j w x "din ke savi pasand e ar- hdh e zaman W ” w u w kortg dvar o kon- got i kon d k eager zun "Do not become like me a possessor of skill in science and art, that you may not like me become despised by the great. (If) you want to become a favourite of the masters of the age, (then) bring a jaunty catamite, play the kenger (o: a cheap two-stringed lute) and beg shamelessly like a kongor (o: a type of mendicant who used to beg and extort money by very noisy and vulgar means). (I.e. Learning and art are of no avail if you want to gain the favour of the rulers of our degenerate age: instead you must serve as a pimp and a buffoon)". ^Obeid e Zakani (d. 1371) 170 THE RHYTHM $257 ei rafte be con - — — V * mjrov o ra - — — ^ - A: a" Aras At j tor a ~ “ V y “ <7 danad u + -i * u u “ ,v" midov o m \j — ajkanad an - iv 1 w — danad it ■■ hamcu " git hie' magu ~~ w w — dar tag o pit — v_y w — danad '/7 s ■ 8 My reason for including this w/wVi among the examples is that its last contains on!) long syllables. Vfy intention was to show that this is not only theoretically possible* but also does occur in the literature (cf El well- Sutton, p, 135). However the example above is not ideal for this purpose, as some editors prefer the reading > b fh j b x b ) h > b f I ft danad o ft danad o ft danad o ft ■i' ■ ■ Another example of a rahaSi-mesra** containing only long syllables may be seen in the following chronogram on the accession ofOurangzib (Classical and Indo-Persian Aurangzeb) by *iil j'lJ -l** Mohammad Lesano-IIah alias ^ Moila Sah (d< ca* 1660), sohh i del e man con go/ e A '^r* sid ' sekoft w w ■ V V ■ — w — haq zaiuT sod qohdr e ha* tid ru raft ta'rix e joiii' x e sah e ou- ranq *’ mar a ■“ W \J “ Kt “ “ ** — zef/odhaq gap [ alhaq 7" ra Iraq gofi "One morning my heart blossomed out like the flower of the sun. Truth was revealed, swept away the dust of futility, (and) told me the chronogram of the Lord of the Throne: “Shadow of God". Indeed, it came true (lit. by Truth 3 : by God, this it said truly)", (Ji: 900 + J : 30 + 1:1 + J: 30 + C : 8 + : IOO = 1069 ATI, corresponding to 1658/59 A.D. Cf, A.K,S. Lambton: Persian Grammar , (Cambridge 1953) pp, 259, 260 and 256), §258 TH E ROB Ad M ETR E 171 "O you who have gone forth like a ball in (i.e. from) Fate's polostiek. Keep walking and running left and right, and do not complain (lit. say nothing)! For He who threw you into this race-and-run, He knows. He knows. He knows. He!” (ascribed to) jt- c Omar Xayyam bar Sesq e loam' 4 na stibr’ pet- da s" na del “ ^ ~ hi ru ve loam 10 * fe> * w — /" qam ke mar a — \j w “ i" del ke tom na Saql ' bar jd s” na del v ^ ^ s u kith e qd- \s \J ” fast' na qam w w ' s” sang e xd~ ” u v — m na del — ^ w u ■“ ‘in the love of you I have neither patience nor heart; without (seeing) your face neither my reason nor heart is right. This sorrow which I have is the mountain of Qaf". not a sorrow. This heart which you have is a flint stone, not a heart”. Rudaki (d. 940) §258 G. Lazard 12 quotes the above roba^i by Rudaki and compares it with the well-known archaic verse (wrongly ascribed to y) Abu Hafs e Soqdf) dhu ye kuhi vdr ■' naddrad dar dost' ce gune davadd bi ydr-' ce gune ravadd 1 3 "How (does it happen that) the mountain deer is running in the plain? He has no companion. How does he go without a companion”? 9 -am goes syntactically with sabr and Jet. 10 -am goes syntactically with Saqt and del , 1 1 A fabulous mountain encompassing the world, 12 Loc. at 13 This is the oldest version, see a! -mo p. 171, 172 THE RHYTHM §258 Lazard points out that the syllabic structure of this verse and the above robust of Rudaki is very similar: bar V esq e foam 1 na sabr peidd st na del ahu ye kuhi I dar das I ce gune ravadd He argues con vinci ngly that the rohd^t metre is based upon a reinterpretation of an accentual Pahlavi metre of which the ahu ye kuhi is another remnant. §259 For all the ouzan mentioned so far. particulars have been given concerning their frequency, but about the rohaSi so much has been written already, that I shall make an exception It is scarcely necessary to inform the reader that this vazn is among the most common, that SOmar Xayyiim wrote poetry in this rhythm etc. In gonj e mxan rhe chapter on almost every poet ends with a selection of his rahaSiau indicated by the title hi rancho "quatrains". §260 This rhythm ts hardly used for anything but rohd^idL i.e. short poems consisting of two abydi with the rhyme arrangement aaxa or sometimes aaaa. One does sometimes come across a single belt, a so-called j fard "single" following this rhythm. An instance is seen in § 156 first example. Poems of more than two abydi in vazn e robdSi arc extremely rare. The following well-known anonymous 14 poem which comprises three abydi follows the rohdSi rhythm increased by one syllable: har kas ke bedd - ash e xerad a z bar kas ke nada- langa " xarak e har kas ke rsadd- nad d 1 5 bedd- gombad e gar- nad o 1 5 bedd - .nr 1 be man - ■i # nad o 1 5 nadd- nad kebeddnad du n bejahdnad nad ke nadanad zel beresdnad nad ke nadanad 14 By some ascribed to Nasir od-Din e Tusi. 15 Though y is elsewhere always pronounced va/er "the abundant metre" (basic rokn: niafakdalon ^ „ w and the three compound metres: /. hahr e lav'd "the long metre" {Jd^uhm v - - alternating with mafaSilon w - - >. xo> ^ hahr e madid "the extended metre" {fit Wdfon - „ - - alternatine with fa Won „ ). Jb . / hahr e hash "the simple metre" {mostatWtm -~ v - alternating with fa Won - §262 One does now and then come across a Persian poem written in one or other oi the above bohur, but they arc extremely rare, so one example must suffice. I he following two abyd t by the Babi poetess 0^' Qorrat ol-^Ein (d. 1852) are in hahr e kdmel e mosamrmn e sdlem "the sound eightfold complete metre ". * t* to wo mo Ik o jd- h e sekandari agar a H x"os as - /•’ to darx "ori man o V esq e d" mah e xub 'ru be nasal o qah - qahe sod ford man o rasm o rd - h e qalandari v agar i“ bad as - /*’ mard sera ke co sod said ye bald bar d ke ana s-sahi- do be karbald U U “ y “ W — W — | V_/ u u - u ^ "You and (i.e. you can have) the power and grandeur of Alexander! Me and (i.e. I prefer) the habit and way of the mendicants! If that be good, you are worthy, if this be bad, (still it is) all right for me. Me and (i.e. I have chosen) $263 STRESS 175 the love of that fair-faced “moon", who when the call to calamity came to him, went down with joy and laughter, saying (lit, that) ‘l am the martyr at Karbala' * * * §263 One of the chief characteristics of Classical Persian poetry is its extreme regularity. One may ask how it is possible for a Persian poet to write a masnavi of thousands and thousands of completely regular lines without becoming monotonous: 'Regularity', in the sense of each foot being exactly like the next and each line being exactly like the next, would be not a mem hut a defect in poetry, ...a series of completely 'regular 1 lines would be, not a proof of poetic skill, but unbearably monotonous * , The answer, at least for Persian poetry as it is recited today 2 , is that, whereas the prosody is based upon the regular succession of long and short syllables, the language has a strong stress accent, which has nothing to do with the verse rhythm. The stress pattern therefore changes from line to line and prevents monotony. Classical Greek poetry shows a similar combination of a regular Quantitative rhythm combined with a shifting accent pattern, probably a pitch accent and not a stress accent 3 . The accent of Classical Persian may have been a pitch-accent also, for languages with a phonemically relevant dis- tinction between long and short vowels usually have pitch-accents, not stress-accents. Modern Persian is no exception to this rule, for as we have seen above (§3), the distinction of short and long vowels though observed in recitation of poetry is no longer phonemically relevant in Persian. * * * 1 Marjorie Boulton: The Anatomy of Poetry. London 1953, p. 29 (in the 1977 reprint). 2 Attempts have been made to analyse the stress patterns of Classical Persian poets, but since we have no certain knowledge of the position of stress in the Classical age, such attempts are futile. 3 ~£he sound of Homer s songs when recited with due regard to both quantity and accent is wonderful. Most scholars, however, while reading Homer, disregard the accent completely and substitute stressed for long and unstressed for short syllables producing very unpoetic results. 176 THE RHYTHM §264 Except for masnaviat all the poems of a Persian poet are usually collected into what is called a divan *. The divan is usually subdivided into sections, each section comprising poems written in one genre: qasa'ed , qazaliiu etc. Within each section the poems are arranged alphabetically according to the last letter of the rhyme. This arrangement is in one respect very convenient, for if we remember or hear a single line by some poet and want to know the rest of the poem we can easily find it. Similarly, if we want to check whether a line of a poem of unknown authorship is by such and such a poet, it is easily done. §265 In the case of a large divan which may have scores or even hundreds of poems ending with the same letter, it may be laborious to search out a line. However, the poems ending with the same letter are usually arranged according to some system. In many cases the alphabetical order is complete, that is poems with the same end letter are arranged according to their penultimate letter, poems in which the rhyming element consists of only one letter coming before the rest. Poems with common penultimate letters are then arranged according to the third last letter etc. This is the way in which the poems of SaSdi are arranged in the edition quoted in §175 and passim. §266 Slightly less satisfactory is the system followed e g. in Dr. Tafazzoli's edition of divan c Altar, 'eheran l%7, where the poems are arranged in the same alphabetical order, but without considering the rhyming element, the place of each poem in the divan being determined exclusively by the final letters of the second mesrSl. §267 In many cases the arrangement of poems in a divan takes into consideration only the last letter o; the rhyming element. This means that poems ending f.ex. in nun are grouped together in one group, and poems ending in vav are grouped together in another group etc., but within these groups there is no further alphabetical arrangement of the poems which may number hundreds in Moulavfs divan there are 145 poems ending in vav and 345 poems ending in nun . However, there is usually some system of arrangement, and if one has to work with a voluminous divan one is well-advised to work out the system for oneself. The system employed in Dr, Foruzanfar’s edition of MoulavPs divan {divan e kabir or kolliat e sams ) is as follows: 4 A few poets like Amir Xosrov e Dehlavi have more than one divan. §267 THE ARRANGEMENT Of- A OlVAN 177 Firstly — as in any other divan — the poems are arranged alphabetically in large groups according to the last letter of the rhyming element. Each of these groups is then subdivided into three groups, the first containing Persian poems, the second bilingual poems, molamma^at, and the third Arabic poems. ■ Each of these groups is then subdivided on a rhythmical basis so that all poems belonging to one and the same hahr are grouped together. The order of the hohur is everywhere the same, namely: hahr e rajaz hahr e madid hahr e hazaj hahr e kdmel hahr e ramal hahr e has it hahr e tavil 5 6 hahr e motaqareb hahr e vafer 6 hahr e motaddrek hahr e molar e *i hahr e xafif hahr e monsareh hahr e sari V hahr e mojiass syllabic poems 7 Within these groups the poems arc arranged according to their ouzart , poems having the same vazn coming together in one group. ! he ouzan are arranged according to the number of syllables, longer rhythms preceding shorter rhythms, and when two rhythms have the same number of syllables, the sdlem “sound" one precedes the "derived" one, e.g. in hahr e rajaz — v- precedes - v which in turn precedes v -. These groups are not further subdivided, but the individual poems of the group tend to be arranged according to their regularity: the more freely the poet has availed himself of the permitted variations, the further the poem is relegated towards the end of the group, and poems containing a regular transgression against the rules, a so-called da^L- seqte “stumbling" come at the very end of the group. In order to make the system quite clear I have shown it in tabular form in Appendix Three. 5 The position of hahr e tavil is not quite fixed, ef. Appendix Three. 6 There is only one occurrence of this hahr in koUiat e xams. 7 Found only in one Greek molamma PARI THREE ADAPTATION INTO OTHER LANGUAGES "Poetry was one and indivisible, the language in which it was written merely an unimportant accident”. Elias John Wilkinson Gibb XXIV THE ADAPTATION OF CLASSICAL PERSIAN PROSODY TO URDU §268 With the gradual stagnation of the Indo-Persian literary tradition during the later Mughal period* Urdu', a Persianized dialect of Western Hindi* became the principal literary medium of the Indian Moslems as well as of the Persianized Hindus of Northern India. Urdu literature is a direct continuation of Indo-Persian literature. In form and content its standards and ideals are those of Persian. Urdu literature is therefore Iranian rather than Indian, but inevitably influenced by its Indian environment. Urdu poetry is Persian poetry with a difference. The Persian scholar who takes up the study of Classical Urdu poetry will find that a new dimension is added to his literary experience. Urdu poetry has repeatedly been condemned as a servile imitation of its Persian models, mere "verbal ingenuity and prosodic dexterity" (J.C. Ghosh) without the “faintest flavour of originality” (Sir Charles Lyall). However* these scathing judgments have mostly been passed by indologists who were prevented by their insufficient knowledge of the Persian tradition from appreciating the achievement of the Urdu poets. §269 Urdu prosody is based on Persian prosody with only such changes as are required by the different phonetic structure of the two languages. A complete description of Urdu prosody on the lines of the description of Persian prosody given in Part One and Part Two is therefore unnecessary. 1 j±j l urdu is derived from Turkish ordu “camp, army” (whence also English horde), and is generally explained as originally indicating the mixture of Persian and Hindi spoken in the army, the ^camp-language" as opposed to the refined Persian language of the Mughal court. This explanation is far off the mark. The language now termed urdu was originally called Hindi “Indian * as opposed to tarsi “Persian ". The word urdu is an abbreviation 4jh of . Jjjj 1 zaban e urdu e muW/a “ihe language of The Exalted Camp ti e. of The Imperial Court)", "The King's Urdu" would be a better translation than "camp-language". 182 URDU PROSODY $269 This chapter describes only the differences between the Persian and the Urdu systems. However, read in conjunction with Part One and Part Two above, this chapter still furnishes a complete description of Classical Urdu prosody. In order to facilitate reference the chapter follows as closely as possible the order of the exposition of Classical Persian prosody. It could be objected that this method makes the work inaccessible to those Urdu students who do not know Persian. However, since no serious study of Urdu poetry is possible without an adequate knowledge of Persian, such students should be advised to take up the study of Persian and the sooner the better. [Ad Chapter /] §270 The adaptation of Classical Persian prosody to Urdu was greatly facilitated by the almost identical vowel systems of the two languages. Like Classical Persian. Urdu has three short vowels: a , /, «, five long vowels d, /, zi, e , d, and two diphthongs ai and an. (In word final position i and e are distinguished in the Urdu script: ^/p larki “girl ”, larke “boys”.) All the Urdu vowels also occur nasalized. In §86 above we saw that long nasalized vowels probably existed in Classical Persian as well. The nasalized Persian vowels, however, must be phonemically analysed as long vowels + fnf, whereas the nasalized Urdu vowels are phonemically distinct entities. §271 Unlike Persian, the Classical Urdu vowel system continues practically unchanged to the present day. In Modern Urdu a and d arc pronounced [a] (or [ a ]) and [a:] respectively. Thus, in addition to the distinction of quantity there is a distinction of quality, ai and au are pronounced [ai] and (ao>] and by some speakers [e:] and (o:J. This, however, does not affect the vowel system as such. Moreover it cannot be proved that these pronunciations are not as old as the oldest Urdu poems. §272 Modern Urdu also possesses three short vowels of ambivalent phonemic status : e, o , ee. At first these vowels appear to be allophones of /, «, and a respectively. / and u become e and o before H + consonant, i, «, and a become e t $274 VOWEL AND CONSONANT SYSTEM 183 o. and (€ before h + consonant. Thus ^ jmuhrj “seal" is realized as mohr (or muhar 2 ), and ^ / mihr / “kindness" is realized as mehr (or mihar 2 or mihir 2 ). jnahrj “canal" is realized as mehr or in the spoken language * more commonly nahar [nahar]-, whereas yjr jnahrul "living by the canal (adj.)". "Nehru (proper name)" is always nahru. d h , 4 ^- e‘\ 4 ^ j\ etc, h as a 2 With an anaptyctic vowel, sec §§282-283. 3 In popular speech 5 disappears with compensatory lengthening: f/o :1a :J. [fe:l], similarly [ma:lu:mj for mu'Uuni “known". Occasionally ham: a too undergoes a similar development: jmu’ntini |mo:min] "faithful, believer", 4 But never nehar for Inahrl. In other words, the development a > (t > e does not occur before an anaptyctic vowel. 5 To these may be added the extremely rare i, by practically all speakers confused with / or r. 6 Pronounced with the tongue “curled hack, slightly, almost as when you try to prevent an elusive pill from rolling off the tip of your tongue" (Bailey, Firth and Fiarley: Teach Yourself Urdu, London 1956. p. XVII. As a self instructor in Urdu this book is not a success, but its introduction to the pronunciation of Urdu is admirable). 184 URDU PROSODY §274 separate consonant is written with the "pendant he latkan vali he, for example Jr. pahal "side”, but Jfc p'al "fruit”. kah "speak", ki "that” jfc*e.g. lak h "100,000". Retroflex consonants are indicated with a diacritical mark ^ O /, i d, j r, r\ etc. [Ad Chapter //] §275 Transcription of the Urdu vowels: jT ■ * 4> — a the same metrically lengthened: a: * t 1 * = ae (see § 272) * i 1 « — i (see §276) < fe fl ■ — u (see §276) ■ < • * l » » » » » » i: fe « ■ & e (see §272) * m ■ -v u » )> » ft » ft u: # * fe * — o (see §272) r* t.. — a » » » » shortened : a csv * — I » » » » » ft m l — e ft » ft » » ft e — u » n » » ft u ^ * ■ ■ 0 » » » » » ft 6 ai ft » ft » » ft ai Sk. _ A « ■ ■) au _ _ _ ft » » » ft au Nasalization is rendered by E.g. a , a , di. For vowels followed by \ see §§ 307-3 1 0. It will be noticed that ’ denotes metrically shortened long vowels, whereas originally short vowels are unmarked. The sign ; denotes metrically lengthened short vowels. Originally long vowels are indicated by ' . We have to distinguish between e.g. /. /, i and §276 We often find Urdu / corresponding to Persian and Arabic a. E g. kisti "boat" from Persian kasti, vidd V "farewell" from Arabic wadd *i. This pronunciation was probably taken over with the words from East or South Persian dialects 7 . 7 Sec r.cx. L.N. Kiseleva and V.l. Mikolayiik : Dan-Russkiy Slovar ", Moskva 1978. s.u. kesti . §278 TRANSCRIPTION 185 kisti and vidaS are considered standard forms in Urdu, but in many other words the forms with i are considered substandard, as f.ex. nimak for ^Ss namak "salt’*. It is therefore often difficult to decide whether to transcribe i or a. Wherever i has been decided upon, it is marked off with a diacritical point: kisti, vidd ¥. Occasionally the Arabic prefix ma~ has become mu - in Urdu, probably influenced by the greater number of loanwords with the prefix mu-. Here also the diacritical point is used. Thus cj: muhabbat ““love” from Arabic mahabbat , and muhaUd "quarter of a town” from Arabic mahallat. §277 The transcription of the consonants call for few remarks: Consonants which are not counted in the scansion are indicated by raised letters. Similarly, the aspirated stops are rendered by k‘\ g h etc., and not — as is normally done — by k/t, gh etc. Classical Persian x" has been allowed to stand, but the various Arabic s‘s and z's etc. are not distinguished {cf. §21). Apart from this I have followed the system used in D.J. Matthews and C. Shackle: An Anthology of Classical Urdu Love Lyrics , London O.U.P., 1972. Accordingly the retroflex letters are marked with a diacritical point: /, i‘\ d, d'\ r, r\ and g (£) is distinguished from q (j). For the rest this system of transcription is identical with that for Persian employed above. [Ad Chapter Ilf] §278 As in Classical Persian the contrast of long and short vowels is a major factor in producing poetic rhythm. Many lines can be scanned following the rules of Classical Persian prosody alone (with the important, but self* evident reservation that k'\ g h etc. must be reckoned as single consonants, as phonetically indeed they are): wedd' etc. Cf, also JUjT kermdn, “name of a town in Southern Iran", from Old Persian karmdna- (See R.G. Kent: Old Persian, p. 180, and cf. E.G Browne: A Literary History of Persia, Vol. I, p. 145). 186 URDU PROSODY §278 vud ' karnd ■ har g h ari us vdr' kd ■w hai vaztfa muj h dil e: ht- mar ’ kd ar 'zu e: casmd e: kau- sar naht tisnalab hu sarhat e: di~ ddr •' kd — \j — — \j — \j — “To long every instant for that friend is the duty of heartsick me. (I have) no desire for the fountain of Kausar (o: name of a river in Paradise). I am thirsting after a draught of (my beloved’s) face s ’\ J j Vali (Rhythm in §§201-203). [Ad Chapter IV\ §279 As in Classical Persian the second major factor in producing poetic rhythm is the contrast of open and closed syllables. However, whereas in Persian a syllable always begins with a single consonant, in Urdu a syllable may begin with two consonants. §280 These consonant clusters are generally word-initial, and the second element is most often y. Both elements of the clusters are distinctly pronounced, but prosodically always count as a single consonant, and we therefore transcribe k\ g\ d 1 " etc. In the Urdu script such consonant clusters cannot be distinguished, even in vowelled texts, as the veneration for the Arab orthography prevents marking ofT the first consonant as vowelless 8 9 . Only context and metre can decide whether to interpret if as kiyd “done, did” or as k'd “what?”. husn e he par- vd kd x "udbi- n o: x'udbra kar diyd k'a kiva mat ne ki izhd- r e: tamannd kar diva •r #* — O' ~~ o “ “ — — “The unselfconscious beauty I have made selfconscious and conceited. What have I done that 1 expressed (my) desire!” Hasrat. 8 didar also means “seeing " and hence “meeting" which is the ordinary sense, and we might therefore interpret the last line as "1 am thirsting to see my beloved’", but “face"’ goes better with ti.inalab “having a thirsty lip". 9 Cf. Sams e Qeis - verdict on initial consonant clusters in § 107 above. §283 OPEN AND CLOSED SYLLABLES 187 Here it is the metre rather than the context which leads us to k'a kiya. With a different metre we would have had to read kiya k>a. (Rhythm in §§199-200). §281 Urdu is a hybrid language created by the addition of a large Persian 10 element to a Western Hindi dialect, which itself contained a considerable admixture of Sanskrit loanwords. The syllable structures admitted in the three constituent elements of Urdu are quite different from one another, but the Persian and the Sanskrit elements are to a large extent adapted to the syllabic structure of Urdu 11 . §282 In w'ord-final position Urdu allows only single consonants or clusters with a nasal as their first element; medially two consonants are allowed 1 '; initially only single consonants or clusters with y as their second element are allowed : 3 . Urdu ordinarily does not allow two short syllables together. In a second short syllable a is always reduced. In the same position i and u are less regularly reduced. Thus the perfect of pakarna - -) “to catch" is pakra (- -), not pakava hahin (^-) “sister" (< Sanskrit bbagini) forms the plurals ^rr. bahine (w^-) and h&hne i4 (--) whence by analogy the new li singulars bahan and (substandard) bcehn 1 5 . §283 Like Urdu Sanskrit does not allow final consonant dusters, but initially triple and medially even quintuple consonant clusters occur. Moreover in Sanskrit loanwords short final vowels disappear creating new final consonant clusters. Final consonant clusters are common in Persian, too. 10 Tfi is Persian dement again contains a large dement of Arabic. In other words, the Arabic dement in Urdu has come via Persian. 11 In the following description "Urdu” indicates that dialect of Western Hindi which forms the basic element of Urdu* 12 Medially two juxtaposed consonants need not always be regarded as a cluster. See below § 288-29 1 . 13 The very rare initial n h and m f * are really aspirated nasals. See the example in §321. 14 For a see § 272* 1 5 The reduced a is treated at length in §§ 288-29 L J hcehn should perhaps be explained rather as the result of a devdopmeht hohin > hafrn > hatm~ Cf. §§307-310 and 271 J 188 URDU PROSODY Thus Sanskrit b'fama- becomes first h h ram and then b k aram “mistake". Sanskrit d h arma - becomes first d h arm and then d h aram "virtue, duty". Persian garni “warm" becomes garam. See also the examples in §272, Before h + consonant the anaptyctic a becomes <\ which may in turn be identified with e. fathpiir “name of a town" thus becomes fatehpur or fatehpur l6 . On the other hand Persian and Sanskrit short vowels are reduced to prevent two short syllables from following one another. Persian (Arabic) harakat “blessing" ^ -) becomes barkat §284 In poetry, however the poets take the utmost care to employ the "correct" forms. Since the Urdu poets were also Persian scholars, but unacquainted with Sanskrit, this means that Persian loanwords in poetry revert to their original form, whereas Sanskrit loanwords are used in their popular form. bkaram k 'id ja~ e zdlim te- re qdmat ki dardzi kd agar us tur~ rd e: pur pe- c o xam kd pe- c o xam nikie W — ™ W — “ “O tyrant, the mistake about the tallness of your figure will be rectified if the curls and twists of your hair full of curls and twists are straightened out (for then your beautiful tall figure will appear short in comparison with your long hair)". Galib (Rhythm in §§223-225). muj h dil ke kabutar kdu ye kdnr ’ d h aram kd hai * pakrd hai ten lat ne tuk us kd ctyrdti jd ■*“ ““ “ | — “Your locks have caught the dove of my heart. Is this an act of virtue? Just release it”. Vail (Rhythm in §§223-225). !6 For - pur see $318, footnote 47. §285 OPEN AND CLOSED SYLLABLES 189 The discrepancy between the pronunciation of Persian (and Arabic) loanwords in the spoken language and in the poetic diction is often surprising. Thus rr^subah (^-) "morning" and i sama $ [fcma:] (^--) "candle" of the spoken language revert to subh (- w subh*) and sam ¥ (- ^ sam ¥ ’) in poetry, as in the following verse in which SAis bravely faces the hardships and infirmities of old age: rf ai sam ¥* subh * hdu i'dri si rah gai hdi hai rati hdi kis lie ■^P use b’f gu- zdr' de “ \J \j — — \,i — i l _/ — "O candle, the morning is coming, why art thou weeping? Only a little is left. Let us get through with that also". (Rhythm in §§237-238.) [Ad Chapter l 7 ] §285 The rules governing the scansion of long and overlong syllables arc identical in Persian and Urdu with just one exception: The word jj\ aur (and its dialect variant jy haur ) “and, more" may be scanned both aur' ( hour *) -w and dur ( hour ) -. The latter is more common. jale hdi sam- ¥•’ se parva- kahi hdi meh- r* bhi jag me nd aur 1 mat tuj h se kahi vafa bhi hai “The moth is being burnt by the candle and I by you. Does love at all exist in the world or does affection 1 ^ exist at all"? Sauda (Rhythm in §§228-229.) 17 In Arabic and in Persian vafa means "fidelity", and "fidelity" is the meaning given in the Urdu dictionaries. However* in Urdu the word has come to be used merely as a synonym for "love" and "affection", and if vafa at all has the sense of fidelity in Urdu, it is 190 URDU PROSODY §285 bay a us ki sardpd kis- nazakat dur var e: . xuhi laldjdt kd **!■ mane ndz o: lik’d id kai add disfd “How long shall I (continue to) write about her grace and delicacy? (I could go on for ever for) in the Realm of Beauty charm and attraction is revealed everywhere (lit, from head to foot)”. Vali (Rhythm in §§191-193.) ft i w* *! ** nab/ sad- qO quibd madad hoi use a- kau jam at - s ' hat t 1 "' hour cd~ r imam ■ "For Qutb, the sacrifice of the prophet 18 , life is all happiness. The eight and four imams are his succour”. Muhammad-Quli Qutbsah (Rhythm in §§178-183.) §286 In § 37 it was mentioned that in traditional Indian recitation of Persian poetry, the •’ of the overlong syllable is realized as an indistinct vowel. In the recitation of Urdu poetry usage varies. Some do not at least only used about fidelity in love. In the following verse by Vail "The Father of Urdu Poetry" the translation “fidelity" would make no sense: yad' aid hdi taupe jab «i gul c: bd- g e vafd ask ' karte hdi maka go- .«i v dama- it * me a — w — V XA “ *- V “ “ \/ V L "When I remember lhai flower of ihe garden of love, tears come and take up their abode in the corner of my skin (after drenching the whole upper garment)"* (Rhythm in §§21 1 -2 1 2. ) 18 l.e. "the Prophets devoted servant '. This is merely a paraphrase of the poet's name. MuhammadQuII, which means "Muhammad's slave ' (from Turkish kul "slave”). §288 OVERLONG SYLLABLES 191 pronounce others do, especially if it follows a consonant cluster. If the second member of the consonant cluster is S, the •* is always pronounced. This agrees with the genius of the language, where a reduced vowel, a "minimally realized a'' is often heard after consonant clusters' 1 *. [Ad Chapter Vl\ §287 The rules for hiatus or hamza (hamze) are the same in Urdu and Persian prosody. However, whereas in Persian relative scarcity of hamza is one of the factors that go to make a poem ravdn “fluent” (§50), in Urdu this is definitely not the case. On the contrary here liaison is much less common than hamza. §288 The reasons for this must be sought in the speech-habits of Urdu and Hindi speakers. In the Hindi script final 20 consonants are written as if followed by a short a, the so-called “inherent a". As to when and how this “inherent a" should be pronounced, the grammarians of Hindi 21 are very vague. According to McGregor "the consonant characters have come to have not only syllabic, but also purely consonantal values”. Phis is because “o has become or tended to become mute finally, ...” His "transliteration indicates whether a given ‘inherent' a in a script form represents a normally pronounced vowel or not ...” With regard to forms like pakra (§282 above) he says that the inherent a between k and r "is not pronounced ... or is only minimally pronounced” 22 . In other words, McGregor, perhaps the most competent grammarian of modern Hindi, does not want to commit himself on its pronunciation. Scholberg gets around the problem by stating that ek “one” “really contains two syllables" 2 \ (e-&). In his rather unscientific, but nevertheless 19 CP, R S McGregor: Outline of Hindi Grammar, p. xxvii, London O.U.P. 1972, 20 Final in word or syllable. 21 Grammarians of Urdu have not paid any attention to this problem, which the conveniently ambiguous Arabic script has camouflaged, as it were, 22 McGregor, op. cit. p. xxiv (italics mine), 23 Scholberg: Concise Grammar of the Hindi Language* § 13, London O.U.P., 1940. 192 URDU PROSODY §288 very useful Hindi Grammar 2 *, Edwin Greaves appears to me to be closest to the truth. According to him " a is understood (and sounded) in al! cases where a consonant bears no other vowel ... In many cases, the a-sound is very slight, but, with care, can be distinguished”, §289 1 would tentatively describe and explain the occurrence in Hindi and Urdu of the reduced a — so I prefer to call it— as follows: All consonants that are neither followed by any other vowel nor are themselves the first member of a consonant cluster 25 — in native words restricted to doubled consonants and clusters having nasals as their first or y as their second element— are followed by a phoneme /al, which behaves very much like the French e muet ; i.e. it is realized only to the extent necessary in order to produce the neighbouring phonemes. Thus it is fully pronounced after an initial consonant (or consonant cluster), before a consonant cluster, and before a consonant followed by reduced a. In all other cases — most conspicuously in word final position- the a is reduced, When several /a/’s follow one another, then— as with the French e muet 26 - every second jaj is reduced: jbaragadaj bar gad “banyan tree”, in Hindi written ef77T jmtaj ant “end' , but (banatd) j hanatd > bunta “becomes”. In the Urdu scrip! only doubled consonants can be similarly distinguished: US' (kl‘} fkuttdi kutta “dog”, but tits' ik'tt') ikatata! hand "spins". The phonemic relevance of the reduced jaf in kdttd is borne out by the fact that in kdttd each r is pronounced with one implosion and one explosion, whereas in kutta we find tl pronounced with an implosion, a pause, and an explosion. 26 Cf. French je ne me te rappelle pas [janmalrapclpoj. §291 HIATUS OR HAM/. A 193 §290 The extent to which fa! is reduced depends upon the speakers personal speech habits (his dialect background), upon the speed and style of speech —the vowels being less reduced in slow and dignified speech -and also upon the phonetic environment of the reduced (af. The Urdu and Hindi unaspirated stop consonants are not like “unaspirated" stop consonants in many languages (e.g. Persian) slightly aspirated, but they arc, as in French 27 , completely unaspirated. Unless followed by a vowel such consonants are practically inaudible. They are therefore in Urdu and Hindi followed by just enough of an ->-sound to make the consonant audible, i.e. they are followed by a reduced jaj. In final position the reduced jaj is “very slight, but, with care, can be distinguished". After voiced aspirated consonants the reduced ja, is slightly more distinct, this time not to make the consonant audible, but to make it pronouncable. Medially the reduced jaj is still very slight, but now easily discernible, especially if it has been reduced as result of the addition of an inflexional ending, e.g. pak"ra “caught" from pakamd "to catch" (cf. §282). In order to facilitate the pronunciation, jaj following cumbersome consonant clusters (occurring only in loanwords) is only slightly reduced and may even be fully sounded, but in ordinary speech the clusters are more often broken up into single consonants, thus d'amv or dharam "virtue, duty", garni" or garam “warm" (cf. §283). §291 The existence of the reduced jai probably has a restraining effect upon liaison between words and is the reason that ham: a is more common in Urdu poetry than hamze in Persian poetry. At the same time this quality of Urdu poetry seems to prove the correctness of the above description of reduced / aj in Urdu and Hindi. In this work the reduced jaj is not shown in the transcription (unless it occurs in an overlong syllable, of course, in which case it is transcribed ’). 27 Is u possibly the presence in hoih languages oT completely unaspirated consonants that account for the similar behaviour of c muci and reduced ia, 7 (In French too the stop consonants are inaudible unless followed by a vowel. Cf. pc fit fpsti). pettic [patit 1 ].) Or docs the reduced tat rc fleet a Dravidian substratum? In Proto-Dravidian "a non-morphemic (so-called cnunciative) -tu follows all stops in final position. The phonetic value ot this final -/u seems to have been a sort of vowel-like release of the final plosive". Kamil Zvefehil : Comparative Dravidian Phonology, p, 35. The Hague 1970. 194 URDU PROSODY §292 [Ad Chapter VII\ §292 The rules for scanning vdv e fa// and izdfd (Persian ezdfe) are the same as in Persian. Surprisingly enough they are pronounced a (or o:) and e (or e:) as in Modern Persian, not u (u:) and / (/. ) as in Classical Persian. Note that unlike Persian there is no y- or u -glide after vowels. In such cases they are written with hamza,^. and >. See the examples in §§278 (second verse), 313 (first verse). Even though they occur only in Persian constructions, they are encountered very frequently. [Ad Chapter VIII] §293 In native words the only short final vowel to occur is jaj which is always reduced, except in the monosyllables na “no, not”, <$>- c h a "six" (also and more commonly pronounced c h e), and *j pa “on, but”. Reduced vowels arc not reckoned in the Urdu prosody (except of course in overlong syllables), na , c h a and pa are always short, but all three have prosodically long doublets t na, c h eh or c h e. ^ par. §294 In Sanskrit loanwords final short i , u and a are always reduced. By educated Hindi speakers < mati > “meaning” is pronounced differently from JTd < mata ) “religion": mat* and mat 0 respectively. In Urdu the two are spelled and pronounced alike mat. §295 The final vowel of Persian loanwords ending in -e (Classical Persian -a) has in Urdu been identified with ~d in native words (e.g. haccd “child” which is declined exactly like the native word iTpl lavkd “boy”) and will therefore be treated together with other words ending in long vowels 28 (§§298-310). Persian na “no, not” is not distinguished from native na with which it is identical in form, meaning and origin. In varna “else” the final a may as in Persian be scanned long or short as the metre requires. In the spoken 28 Some grammarians and lexicographers distinguish between original Persian -a and indigenous *d, thus Bailey who writes face a* but te[ka^ But there is not — and judging from the prosody never was* any corresponding distinction in pronunciation. In the Urdu script the spellings with he for original Persian - a and ahf for indigenous are often confused. §296 FINAL SHORT VOWl LS 195 language it is generally pronounced varnd, but also varna 2 '\ and I have even heard varan [varan], but not varn". In Hindi it is spelled < varand) or <37^7 < vara no) . In ki 29 “that *, tdki “so that*’, agarci “even though". «5cVU- hdlaki “whereas", etc., / is generally scanned short, but may also as in Persian be scanned long. With some speakers ki “that " and S ki “of (f.)“ form a minimal pair, ki being pronounced with a hypershort, (but not reduced) i. Others - probably the majority -do not distinguish in pronunciation between ki and ki. [Ad Chapter IX] §296 In Urdu there is a phonemically relevant distinction between nasalized long vowels and long vowels followed by n : uU md means “mother”. oU man means “honour”, ma is long, man is overlong. In Persian there is no such distinction. Persian (and Arabic) loanwords with n following a long vowel are therefore allowed a double interpretation mdda “tired" may be scanned mdda ( — ) as well as man 'da (- ^ ^ j^i dsman par “on the sky” may be scanned as'md par (-„--) as well as as'man' par (- w - ^ -). In Persian the latter alternative is extremely rare, but in Urdu quite common. In word final position n is distinguished from nasalization in the script, the dot of 0 being omitted when J indicates nasalization: kard kaj jabi pa sar e: kafan ki gurur e ds~ q ' kd bdk 'pan mere qdtild kd gumd na ho pas e marg ■’ ham ne b h ula diva y y ” y — V U V ’ W U U V V ’ “Place the end of the shroud aslant on my forehead. My murderers should not think that after death we have given up the coquetry caused by taking pride in our Love 30 ”. Faiz (Rhythm in §262) 29 For the phonemic status of final -/ and in Persian loanwords cf S3Q3. 30 Love here stands for (leftist) ideology. Faiz, who wrote this poem in Montgomery Jail, uses the traditional mystical imagery in a modern political context. URDU PROSODY §296 196 mil jae ek' an' ham di- ne 3 1 ke sdm '~ me kasratnu- mdiva w ne jab 'ake hu kare ^ ^ ^ \j \y \j ^ ^ "All appearances of' plurality vanish in one instant when we come before the mirror and say Hu i2 ". >>> Dard (Rhythm in §§237-238) In the first example Arabic jahin and Persian gumdn must be scanned jabt and gumd, but in the second example Arabic an must be scanned an'. In the following two verses by Sauda bagbdn must be scanned bag'bdn’ and bag'bd respectively: rux sat hai jdte hai bag'bdn' vd 3 3 jahd se ki tuk dek h •’ p h ir a yd na le caman jdegd KS — W — ■ — “Do we have the permission, o Gardener, to glance at the garden? (Allow it, for) wc are going there from whence there is no return". (Rhythm in §§237-238) hubdb e: v caman kd w — — lab e; ju tere kd- \j — hai ai bd- i dam de- “ ” g 'bd ham k**te hat 31 Cf. §318. 32 The mystical syllable hu is used for concentrating the thoughts upon God. Originally it is the Arabic third person personal pronoun "he" i.e. "He, God" (Cf § 126). The verse contains a beautiful double image: Firstly, when we concentrate our thoughts upon Him. all else disappears from the mirror of the mind, but also, if we place ourselves before a mirror and say hoo our breath will blur the mirror's surface and all images will disappear. 33 To be pronounced with a breathy vowel, see §§309-310. §299 NASALIZATION 197 “O Gardener, we are (ephemeral like) the bubbles at the edge of the stream. We look at your flowerbed (only) some moments’*. (Rhythm in §§ 173-177) §297 Short nasalized vowels occur only in native words and are short. 6 i l i >. * 1 U jag me koi ki na hasne ( frt ofcji na tuk hasa ft me ro diva hoga hoga V V-- \J ~ ^ “No one in the world has (ever) laughed the least bit without weeping at the same time (lit. that he did not weep in laughing)’*. Dard (Rhythm in 230-232.) [Ad Chapter X] §298 In the treatment of final long vowels Urdu prosody differs most markedly from Persian prosody. In Urdu all final long vowels and diphthongs whether nasalized or not may be shortened as the metre requires. In almost all the Urdu examples quoted so far instances of shortened long finals occur. In the following verse ki and ki , hai and hdi occur side by side: i/jC if* 0 ntd' us ki teri zulfe hai ditndg us ka hdi rate jis ke bdzu par paresa us ki hai ho gat — w ~ W — — — ‘ w — "(Sound) sleep is his, brain (i.e. sanity) is his, (joyous) nights are his, on whose arms your locks have become dishevel led”. Galib (Rhythm in §§ 199-200) §299 The above rule of course applies to Persian loanwords ending in -a from Persian e (Classical Persian -a). See examples in §§278, 284 first verse, 285 first verse. 198 URDU PROSODY §300 §300 The demonstrative-anaphoric pronouns "he, she, it, the, that” and c "he. she. it, the, this" are in careful elocution pronounced vah' and yah 1 respectively. In ordinary speech as well as in poetry they arc pronounced vo and re, which may, of course, be shortened to v<5 and ye when the metre requires, as in the following famous verse by Gaiib. Siiq-' par zo- ki /ague r' rnthi hai na fage ali- ve vo at is ■ r ’ buj'tie / * U Gdfih - na bane vuv - "On love no force (can be applied) for this is the fire, o Gaiib, which when kindled will not burn, and when slaked cannot be kilted (lit. slaking will not succeed". (Rhythm in §§211-212) §301 In native words as well as loanwords of the type mentioned in §299 long final vowels and shortened final vow'els arc equally common. In Persian and Arabic loanwords (including those ending in original long a) the metrical shortening of final long vowels is rare. In the example below we find da Vw for daSva. f * r ~ ~ mere sujv- d' ki dair o: harant se guz- ri qadr rak ’it hu da V- vd (ere dar pa jabhdsd- / ka "The merit of my prostration transcends (the merit gained by prostrating oneself in) temple and mosque. (For) I do maintain that 1 have rubbed my forehead against your threshold". Sauda (Rhythm in §§228-229) §302 The final long vowel of the first member of a compound may also be shortened ( jabhd-sdi in §301), as may a long vowel preceding the future signs -gd, -ge 3 * etc. 34 The reason why long vowels may he shortened before the future sign is that this was originally a separate word. It is derived from Sanskrit gala- “gone ", dikhauga literally meant “1 am gone to show”. In everyday speech, too. long vowels arc shortened before the future sign. §304 FINAL LONG VOWELS 199 /j UI. 4 L £ * * * ':4- £‘)bb dik h dugd tuj h c zdhid us dfiat e: jd kd x (dal dimd- W — — g' me tire w V “ ~ hdi pdr 'sd- i kd s/ — — “ “ "O zealot, I will show you that calamity of the soul (i.e. my beloved, who causes me to neglect and forget religion). There is disorder in your brain — caused by (lit. of) chastity (and a glimpse of my beloved would immediately cure that)'". Sauda (Rhythm in 228-229) §303 T he prosodically ambivalent value of long final vowels reflects their pronunciation in the spoken language, where they are pronounced considerably shorter than long vowels in other positions, often almost as short as short vowels. (This is the reason why English loanwords ending in short vowels are invariably rendered with long final vowels in the Urdu and Hindi scripts. Thus city becomes Jj— siti and isitf)). It would be wrong to conclude that the distinction between long and short final vowels is neutralized in Urdu. However, words like ki "that”, varna "else” (see §295) arc not sufficient to warrant a distinction between long and short final vowels. For a correct description of the Urdu (and Hindi) vowel system we must say that in the first syllable there is a phonemic distinction of long and short vowels. In word internal position we must distinguish between long vowels on the one hand and short or reduced vowels on the other hand. In word final position long vowels contrast with reduced vowels. It is important to notice that the difference between short and reduced vowels is not phonemically relevant. §304 kdi "some” can be scanned in four ways: kdi (see the example in §297), kdi (§296 last example), kdi and kdi jo kdi qil hud mdh'nt kd Saqi ' k h okar dors ' dik'd ? isq ■’ kd Sd- 200 URDU PROSODY $304 "Whoever attended the school of love {lit. saw the lesson of the moon-faced one), lost his wisdom, (yet) was made (lit. became) wise by (lit. of) love". (Rhythm in $§201-203) r>- Siraj Here we have to read koi. An instance of koi is seen in the following verse in which Galib complains over "religious" people prompted by hope of reward rather than love of God. taSat me doza.x me td rahe na dal ■’ do kd- may o: atiga- bt ki lag i lekar hi- hist ' ko "Take and throw Heaven down into Hell, somebody, in order that devotion should not contain (any) love o wine and honey". (Rhythm in §§237-238) §305 A short vowel is often contracted with a following long vowel into a diphtong. Thus gai (^-) and gae (^-) "went” may become gai (-) and gae (-), and hui («^ -) "became" may be read hut (-) as in the following examples : tuf muk" ki par as t is me ai but ki pujanhdri w W — gaj Sumr' meri sari us but kd pujdti jd "My whole life has passed in worshipping your face. O idol-worshipper, allow that (other) idol (i.e. yourself) to be worshipped”. ValT (Rhythm in §§223-225) guzar gae di- n* ’ dur dunya se tis par terd g h ar 'aur kai manzil rahd hai “We have passed from (i.e. given up) religion and world (i.e. both our §306 FINAL, LONG VOWfcLS 201 material and spiritual welfare). Still, your house remains several more stages (away)”. Mazhar (Rhythm in §§194-196) mauj e beta- hi e dil as- k' me hid jab vdnumd y jab hasi zul- / e sanam tab- e paresd- tv' me a “The surge of my heart's impatience has become manifest in my tears ever since my beloved’s locks came and took possession of (lit. settled in) my distracted mind”. Vali (Rhythm in §§211-212) See also the example in §307 (tat for tai "to, up to"). §306 Finally in a verse (or halfverse or before a caesura ) two long vowels may be contracted. In the following example we find calcaldo for calcaldo 35 “going and coming", and jay for yae 35 . • * sdqiyd yd it ' jab taiak has w — — lag ruhd hat cal sake sd- \j — — calcaldo gar cale "O cup-bearer, departures (from this world) are taking place (continually, so) let the cup go round as long as only it can". Dard (Rhythm in §§201-203) dilkast cd - sarkasi no- JS£ /•’ me aisi Z’ 1 me aisi — — ki si tare ruk jay ki gavarnar j h uk jay 35 It is perhaps better to analyse calcafao phonemically as catcalar. In that case no special rule is needed to explain how cakaiav can be scanned finally in a verse. Similarly jae might be analysed as jay. 36 16 be pronounced with a breathy vowel, see §§309-310. 202 URDU PROSODY $306 “Such loveliness in her gait that stars stop (in their course)' such pride in her bearing that governors bow". ^pLUll jf 1 Akbar Ilahabadi (Rhythm in $$211-212) §307 The rules staled in $$305 and 306 may even be applied to vowels separated by h. palmed (w^-) “reached" may thus become paused (- ) and naht may become nai h . paused jii dp ' kd to maWunr' ah hud ki mat paused xu- da ke tdj hahut mat hhi dur' thd “When I reached (my) Self, then I reached God. Now I realized (lit. it became known) that 1 too had been (lit. was) very far away (from God when searching Him in mosques and temples)". ^ Mir (Rhythm in §§237-238) §308 When such contractions occur hamza is occasionally substituted for he in the script. In the following example we find na] 37 for usr naht to “not". (Instances of ndi h — and in the same verse naht too — may be seen in $321). G dqihat ho- egd k'd ma f- litm * no] 4 7 dil hud hat muhtald dil - ddr ' kd — xj — — — “w — “1 don't know what the end is going to be. My heart has become (too) involved with my beloved (to care for the consequences)". Vail (Rhythm in §§201-203) §309 When the vowels preceding and following he are both d. the resulting contraction is in the script rendered only by a long d, thus we find very commonly ul yd and of* w? for yaha “here" and jlj valid “there”. * ^ * See §§296 third example (vo) and 306 first example (yd). 37 To be pronounced naiK see §310. §311 I INAL LONG VOWELS 203 §310 These somewhat surprising contractions of vowels across an intervening h are in complete agreement with the speech habits of Northern India. In Urdu and Hindi the phoneme //;/ is in slow speech realized as a glottal fricative, usually voiced ffi). In fast speech }h( is realized as the supra segmental feature "breathiness". In other words [ft] disappears and the adjacent vowels are instead pronounced with a breathy or //-coloured voice quality 38 . Whether written ^ or yy . the contracted form o' nahi must be pronounced i r breathily nai h . Jij and jl should be pronounced vd h and va h . In the script this pronunciation is very rarely shown with the spellings jUj and oL|_, which very aptly (and much better than my transcription) render the breathiness. [Ad Chapter Xf\ §311 As mentioned in §§283-284 Persian (and Arabic) loanwords are in the spoken language adapted to the syllabic structure of Urdu, but revert to their original form in poetry. Thus Persian norm "soft" and Arabic rizavt “Rizwee (a surname)" become naram and rizvi in the spoken language, but revert to narm and rizavi in poetry As might be expected popular forms with anaplyctic vowels or suppressed a's turn up here and there in poetry. Instances are, however, surprisingly rare. In the following two examples we find tarah for tarh "manner" and qadmd for qadamd "feet”. hamsuxan ti- se ne Jarful- d' kd sir! st* kiya jis tarah kd ki kisi me ho katndl ac - c‘‘d hai — ^ — y u “ ^ u y “The mattock made Farhad acquainted with Sirin. Whatever kind of excellence one may have, it is good". Galib 38 Bailey. Firth & Harley: Teach Yourself Urdu, pp. XXVHI-XXXII (London 1956) teaches how to pronounce hreathy vowels. A thought provoking discussion of the relationship between segmental phonemes and suprasegmental features is given by J.R Firth (who uses the word "prosody” in the sense of suprasegmental feature) in his article Sounds and Prosodies. Transactions of the Philological Society 1948, pp. 121-152. (London 1949). 204 URDU PROSODY 31 1 arz' ki mat ne ki ai gut- ciaulat o: V iz - zat o ima (y (zM^cS .<«/; e fitrat fere qadmb ki bahar pa nisdr — \j “ w w “ — u v — “Humbly I said: ‘O glory of nature's rose-garden, at thy feet (1 would fain) offer my wealth and honour and faith ", Akbar llahubadl (Rhythm of both examples in §§21 1-212) §312 Doubling of consonants for metre’s sake is very rare in Urdu. In the verse below we must read uti h d for til'd "rise!" * ■■ * ink'd me- ■ * kdx e: u- ri 39 dun yd kc J garibb kb mara ke da - r o dtvar ' jagd do hitd do — kj — — “Rise, wake up the poor ones of My world! Shake the door and wall of the palace of the rich"! JL51 Iqbal (Rhythm in §§220-222) §313 In Urdu prosody / is occasionally changed to y before a vowel. Thus rakHo ( v v -) “place!" may become rak'yd (- -). For kijie (- w -) “do" ! we often find kije (- -). Here kije (Prakrit kijjaf) is the older form which has survived in the poetic tradition beside kijie « a modern form showing normalization of the ending -/a kahie A0 \j — w “ “Envy is the punishment for perfection of words (o: my perfect poetry). What to do? Spite is the price of art’s merchandise. What to say"? Galib 39 See §§315-317. 40 Or koine. §315 ALTERING THE SYLLABIC STRUCTURE 205 samaf'ke rak h - ki us nava- yd qadam das- h ' me sauda t e xar' me hamhndpd niajnun 4 1 h h i hat “Place your feet discreetly in the plain of thorns (o: love), O Majnun, for Sauda is (competing with you) in that Held, (and) barefooted too"! (Rhythm of both examples in §§228*229) Sauda §314 The opposite change — the metrical change of y to /—is extremely rare. In the following line by Soz we have to scan \S k'd “what"? as kid (against §280). sdz ■’ k'd (i- yd * 1 adam kd c'dr'kar dun- yd me tu vd A2 tuf'e k y a t h i kami tuj h kd kid dar - kdr ' t'd - — “ V — “* v — “O Soz, why have you given up (the way leading to) annihilation 4 J (in God) and come into the world? What want did you suffer there, what need" ? (Rhythm in §§199-200) [Ad Chapter XII] §315 What has been said about poetic licence in Persian poetry in Chapter XII applies mutdtis mutandis to Urdu loo. However, the Urdu poets seem to use their poetic licence less than their Persian models, obviously because there are more rhythmical ambiguities in the language itself to be exploited. (Cf, in this connection especially §298 above.) 41 LailiTs ill-rated lover who represents the sum total of desperate love. 42 Pronounce vah (cf, §§309-310), 43 Again a pun: lactam also means “need" and '"want". 206 URDU PROSODY §316 §316 The commonest poetic licence is the shortening of e to e in some very frequent everyday words. u« mercl ‘'my”, ^ tera "thy” and ek "one” becoming y mera \j tera and ^ ek. In the first example below we find teri and teri side by side, in the second example ek and ek. tamanna teri a- hat teri r'zu hat agar hai agar V/- tamanna r 'zu hai ’ \J — “ 'w' JW— \J ‘The wish is yours if there is a wish. Yours is the desire if there is a desire 44 ” Dard (Rhythm in §§173-177) majlis me rat ’ ek 1 ' k'a sam ?■’ k-a pa tang' tere parta - v e: hagair har ek he hu- zur' t f, a * * W W ^ V O *** \_/ “-* "Lacking solely thy splendour (o: when thou alone wert missing) at night in the assembly, everyone was disturbed 45 , even candle and moth”. Mir (Rhythm in §§237-238) §317 Because of their sense and employment words like “my", “thy” and "one” are prone to shortening 46 . There is therefore nothing surprising in the shortening of mera , rent and ek to mera , tera and ek. In dictionaries and elsewhere the shortened forms are often transcribed. mira t tira , ik, still they must be pronounced with a short e. This is borne out by the fact that wc occasionally find mera , tera and ek with long e 44 This clever double entendre may be understood in several ways. The poet seems innocently to say, if there is anything you wish, it is entirely up to you, i.e. 1 shall willingly comply with your wishes, or perhaps, your wishes do not concern me at all, but at the same time the poet is passionately pleading, if I have any desire at all it is you I desire, or perhaps he means to say that he longs for God atone. 45 Untranslatable pun on he huzur "absent” lit. "prcscnceless”. Every one was without thy presence and therefore disturbed and absent in mind. 46 Thus English "one" is shortened to "a” in unstressed position. S 319 POETIC LICENCE 207 in the script, where the metre requires short c. See the example in $312 and cf. the use of f and u to indicate / and u discussed in $318. §318 We sometimes — and in the oldest Urdu poetry very often— find and j used to indicate short i and u. Evidently the writers felt that the vowel quality was more important than the vowel quantity. It is certainly irritating for the reader that words like ul b ar “here" and ud'ar "there” are written alike yoi , so spellings like and yoy have something to commend themselves 47 . When such spellings are met with in poetry the vowels must of course he pronounced short. JUJjiA if-. * cj iju£ f 1 '/ f 1 L tit 1 ' 4 * gn/e kd nid ’ at! cal terd pd- haif ’ yit bet- tithu** dida- tuf'e sfda- i'na f'uld- I 'pan (era mat it 48 mat w 4# mat asgar w "Rise, let me wash off the blood of your throat. Let me pul you to sleep (if you) feel sleepy. Come, let me rock your cradle. Alas (that) such (should be) your childhood, O Ali Asghar”. Hasim SAIT (Rhythm in §§230-232) Notice that in the word dind "mirror” the second syllable is in poetry as in everyday speech normally short. See the second yexample in §296. §319 Imitating as it were the archaic spellings mentioned in the previous paragraph, Urdu poets occasionally allow themselves a rather jarring lengthening of / or u. 47 In the modern orthography ^ and 3 as indicators of short i and u are banned, with just one curious exception, The ending -pur in names of towns is always spelled jj Thus jounpur “town in Eastern U P." and hence also icnmpuri "inhabitant of Jaunpur*\ 48 u here does not indicate a metrically shortened vowel. It indicates a short vowel only graphically long. 208 URDU PROSODY §319 * da rd' kite*' maS- lum' hat ve log ' sab kis taraf se at 1 t h e kb d‘'ar A< * caie Ala$ so , is it at all known, where all these people came from and whither they went"? Dard (Rhythm in §§201-203) §320 Archaic or rather dialect forms are frequent in Urdu poetry. !n the examples given in this chapter we notice kdu 5 1 (§ 284 second example and §285 third example) and ke tat (§307) for kb "to”, mane (§285 second example) for me "in”, taiak (§306 first example) for tak "up to”. w«/ A (§278) for mere "my”, tuj b (§305 First example) for tere "thy”. rak h io (§313 second example) for rak h ie “place!”. §321 Editors and scribes sometimes substitute the standard forms for the dialect forms used by the poets and allowance must be made for this when scanning. In the following example we have to read the dialect formal** m h are for the standard form hamdre given in the text. kbi m h dre dard ' kd malt- ram nahi ds'nd nai h dds nai 1 ' ham ~ dam nahi “I have no one to confide my sorrow to (lit. not any confidential of my pain), no fellow, no friend, no companion". Siraj (Rhythm in §§201-203). 49 Perhaps the transcription ki.d^r should have been preferred. 50 Giving at the same lime the poet's nom de plume. 51 In the age of Vali and QutbSah probably pronounced kau h (< Old Hindi kdhu i cf. §307 and 310.) RHYTHMS 209 §323 [Ad PART TWO) §322 On the whole the metres and rhythms popular with the Persian poets are popular with their Urdu brethren, but there are two noteworthy exceptions. The sarM metre (§§216-219) is very rare in Urdu. On the other hand, the metre described in § 262- — though extremely rare in Persian poetry- - is quite common among Urdu poets. The first example in §2% is in this metre. §323 Occasionally Urdu poets employ more or less adapted hindi “native Indian" rhythms. A description of these falls outside the scope of this work 52 , 52 The basis of most of these rhythms is a foot consisting of four ntorae which may be realized in the following ways: u o o w — ^ y w v SJ w The hindi rhythm most commonly met with is v..v _ | y ^ _ v,v _ Examples may be seen in Matthews and Shackle: An Anthology of Classical Urdu Love Lyrics. London (O.U.P.) 1972, pp. 63, 73 and 135. XXV THF ADAPTATION OF CLASSICAL PERSIAN PROSODY TO KARAKHANIDIC (MIDDLE EAST-TURK1SH) §324 Chapter XXVI will deal with Ottoman Turkish prosody. However, since Ottoman Turkish poetry is heavily loaded with Persian and Arabic loanwords, and moreover these loanwords follow the rules of Classical Persian prosody, it will be useful first to look at a purer Turkish idiom. For this purpose I have selected the Karakhanidic 1 Kutadgu Biiig “Auspicious Knowledge", a book of advice (cf. §181) completed in 1069/70 by Yusuf Ulug Hacib 2 . ■p 4 §325 It has often been stated that the "Perso-Arabian prosodial system ... is essentially unsuitable" to Turkish poetry, “for while the Perso-Arabian prosody is quantitative, there are, strictly speaking, no long vowels in the Turkish language” 3 . This is only a half truth. The two major factors in bringing about poetic rhythm in Persian are the opposition of short and long vowels and the opposition of open and closed syllables 4 . The former plays no role in the Turkish languages, but the opposition between open and closed syllables is found in all of them, and it is therefore possible to compose poetry in accordance with the Persian rules in any Turkish tongue. I An Eys (-Turkish language formerly spoken in Sinkiang around Ka$gar. Examples are quoted and numbered after Retjid Rahmeti Arat: Kutadgu Bilig (Istanbul 1947) and I follow his transcription except that I write n for ng. Note that a dot underneath a vowel indicates that it is not written in the original alphabets (Arabic and Utghur). E.J.W. Gibb: A History of Ottoman Poetry (London 1900-1909) vol. I. p 104 (hereafter referred to as Gibb). M. Fuad Koprulii even asserts that "la structure dc la langue turque etait entierement hostile a ce systeme de versification de type quartliiatif. propre a Vara be". La metrique farii: dans la poesie turque in Phtbfogtae Turcicae Fmdamenta II (Aquts Mattiacis 1964) p. 254. See Chapters III and IV above. §327 OPFN AND CLOSED SYLLABLES 211 §326 Now consider the following lines from Kutadgu Bilig : 326 tuzunlug 5 kerek hem siliglig 5 kerek uku$lug kerek hem biligiig kerek w \J> \j — “He must he gentle and he must be eloquent. He must be intelligent and he must be wise (lit, a gentle one is needed, an eloquent one. too, is needed, etc.)”. Here the rhythmical effect is brought about exclusively by the opposition of open (tii, ke- % si-) and closed (-ziin-, -lug, - rek ) syllables. The metre is motaqdreb { miitekarip ) as we would expect in this type of poetry (cf. § 181). However, such completely regular successions of open and closed syllables are rare and have not been noticed before. Some scholars have held that Kuiadgu Bilig is not at all quantitative, but like “the true spontaneous poetry of the Turkish people" 6 is syllabic, the number of syllables in each line alone bringing about the rhythmical effect. Others rightly state that Kuiadgu Bilig is in the motaqdreb metre, but are unable to give their reasons for this view', saying only that the verses are defective and irregular 7 . But, as we shall see, the rhythm o 1 Kutadgu Bilig is neither defective nor irregular. The reason why this has so long escaped notice, may be that everyone has been busy looking for an opposition of long and short vowels and not paid sufficient attention to the consonants. Two simple rules will reduce the number of rhythmically “defective" syllables to about one per cent and a few additional rules will account for most of the remainder §327 The first and most important rule says that any word boundary may optionally count as a consonant according to the requirements of the metre. Long passages of Kutadgu Bilig can be scanned regularly with the help of 5 Amt wrongly reads tuzunluk and sitigtik . 6 Gibb, vol. I. p. 104, cf. ibid. p. 71. 7 ’Les poetes turc ... emploient ces formes ... dc fa^on si defectucusc que. pour pouvoir adapter les mots ... ils doivent ies alterer dans leur prononciation. un mcmc mot etanl prononce dc fa^ons tres dif’fercntes scion sa position dans di Keren ts vers d un meme poeme”. Kopriilti, op. cit. p. 255. The thesis that the Turks preferred Persian metres of eleven syllables, because their original syllabic poetry — of which we know nothing whatsoever — had eleven syllables, is unnecessary, for the oldest extant Turkish poems are all narrative poems and as such must have eleven syllables according to the Persian rules. Cf. §§ 1 38- ! 39 above. 212 KARAKH ANIDtC PROSODY §327 this rule alone. In the following a prosodically significant word boundary is marked with #. 542 okii$ #ed- gu sozler til in sbz- ledi# i$ih #it- ge men tip kattg kur badi# 543 bu yahhg bolur bu # kisi# #ed- giisi# kifi# #ed- gust# #ol bodun viid- * v ■ gust# 544 negii# tir #e$itgil budun ba$- far er kamug #i$- ni yetru# koriip #i§- ler er 545 kimih dev- m leti# has koti irse# ■ ■ FI #oru# kamug #ed- git kilgu# budunka# It tor u# 546 kimin #el- gi bolsa \# budanka# #uzun silig boi- gu kilki# hi met# tiiziin 547 vorik hoi- ■r * sa k inwin* budunka# sozi# SUfig tut- gu til so: me kodki # #dzi# 548 yayig #ol bu devfet ■ * #iter hem buzar ne #irsel turur ter- k irikse # tezer 549 bu kutka # #manma# #usa# #ed- gu ktl hu kiin mun- da erse# yarm #an- da bil 550 bu kutka # kii venme# #ay 8 9 kut hul- mi# keligli# turur kut m yana # bar- 551 ay 9 dev lei #idisi# 10 bu dev/et i ■< bile# turayin 1 0 Use# tur sen edgii# tile # 552 saha# teg- se beglik #u!ugluk #oka# kieiglik #anuk tut #oriin bol- guka# yj \j \j - “He spoke many good words with his tongue. I shall perform your task (i.e. do what you want me to do), (he said and) buckled tightly his belt. (Of) this type is this “best of men”. The “best of men” (is) he (who is) a carrier of (the burdens of) the people. Listen (to) what he says, the man standing at the head of the people, the man acting (only) after having looked thoroughly at everything: (He) whose fortune raises up its head, must make all good laws for the people. 8 Aral reads kimin and gives kimnift in the critical apparatus, Since kimin does not suit the rhythm, I prefer to read kimnm . Dr. Saadet Qtgatay in her Turk Lehfvleri Omeklcri (Ankara 1963) p. 90 also reads kimnift. A correct knowledge of the prosody of KutaJgu Bilig would he very helpful in establishing the original text. 9 See §328 10 See §329. §329 WORD HOUNDARttS 213 (He) who gains power over the people (lit. whose hand becomes long to the people), pure must be his character, just his acts. (He) whose word is current to (i.e. obeyed by) the people, pleasant shall he keep his word and low his person. * hangeable (isi it, this fortune. It makes and unmakes. How fickle it is! When it turns away it flees swiftly. Do not trust fortune, do good if you can (i.e. as long as fortune enables you to). Know (that) today it stays with this one, tomorrow with that. O you who will find fortune, do not rely upon this fortune. Fortune is (a guest) who comes (and) who leaves again. 0 possessor of fortune, if you say ‘let me abide with fortune', (then) you (must) keep on wishing good. 1 f greatness and kingship come to you, know (then that you must) keep meekness ready for when you are grey”. With the exception of 550 and 551 all the verses above scan regularly. In some verses we find a double # #, This does not mean that we have to reckon with a double word boundary. It is simply a device intended to show how the passage is to be understood prosodically. The sign # may in fact have had a phonetic realization as a slight pause. From the point of view of rhythm, a syllable preceding a pause always gives the impression of a long syllable. §328 In 550 and 551 a “minor” rule is illustrated: The vocative particle ay is prosodically short. Perhaps this is not a rule of prosody, but an indication that this word was pronounced e or i. §329 The second major rule says that a non-final open syllable following another open syllable in the same word is always long. Thus idisi and turaym in 551 must be scanned idisi and turaym respectively. (It is important to note that forms like * idisi and * turaym do not occur.) In the following verse there are four instances of this rule: 1819 kiinme# tenediir turn# #og~ leniir negiini # tilese # bilur #dg- reniir "(The child) keeps growing day by day and acquires reason, (so) whatever he wants he knows and learns”. Even originally short vowels in Arabic loanwords may be lengthened according to this rule: B 12 bu me$rik meliki# ma^inlar begi# biliglig #ukuslug #ajunda # yigi# 214 KARAKHANIDK PROSODY §329 "The ruler of the East, the lord of the Chinese, the wise, the discerning, the best in the world (appreciate Kutadgu Bilig)". §330 If three non-final open syllables follow one another, naturally both the second and the third of these are prosod ically long in accordance with the rule given in the last paragraph: 1 794 a tan #og- return#- di erdem bilig legume- di mindin vime# #dk #elig "Your father (who had died) could not teach (you) virtue and knowledge. From me, too, guidance (lit. reason-knowledge) has not been able to reach (you)". §331 An important minor rule says that originally long vowels in Arabic loanwords are optionally read long or short. In the following example Hacib the author’s title, is read Hacib. (In §333 footnote 1 1 an example of Hacib to be read with short a may be seen). 507 ay it it# bu Hacib sozin vet - rtirek tileki # ne ermi$ tak i# but- riirek "This Hacib asked (him) (lit. with words made say) carefully and thoroughly what his request was", §332 It should be noted that there is no prosodically significant difference between long and overlong syllables in Karakhanidic Turkish. In other words, a sequence of three consonants (like kend tutor in the following example) has metrically the same value as a sequence of two consonants: B 34 yine# bu# kitab kbr kamugka # yarar melikler- ke artuk #i/ig kend tutar "Look again at this book, beneficial to all (and) most (of all) to the kings, the town-holding lords". §333 Scanned according to the rules given above, Kutadgu Bilig has few irregularites. Some of these are undoubtedly due to corrupt readings. A careful prosodical analysis of the complete poem might reveal some additional rules, which would further reduce the number of irregular syllables. In Chapter XII (verses 398-461 in Aral's text) which has a total number of 1408 syllables, we find ten prosodically irregular syllables, all of them irregularly lengthened vowels: 399b yd$t, 401a seviglini , 40 !b kaciglika , 415a dig, 421a ifi, 440a ijine, 440b yafina, 443b iki, 456b menu , 458 favi. In 456b, 401a and 401b we are probably confronted with corrupt readings. $333 PROSODICALLY LONG VOWLLS 215 In the critical apparatus we find the prosodically acceptable readings, meiigu (456b) and ka<;tghgka (401b). We may then conjecture sevigligni in 401a. For the other irregular syllables no certain explanation can be given 11 . 11 1 am aware of the danger of proving nothing by attempting to prove too much, and therefore add the following reflections on (he irregularly lengthened short vowels with reluctance and reserve It seems that these irregular lengthenings take place only in a certain limited number of words. Among the seven instances quoted there are two occurrences of both ray and iy. We find cm" again in 462b. dig is very common e g. 1764. 1802. 1805. 1815 and (he following example: 505 dtgke# yaktn Has Hacih #cr- di hir ay 10 #er* sig i/ig bile# so:- kv sir “Close to the king was one Has Hucib (having! secrecy for word with (i.e. who was a confidential adviser to) the most manly king". IT the occurrence of irregularly lengthened vowels is really lexically conditioned, we must conclude that these vowels were long in an earlier stage of the language, and this again would be a clear proof that Kutadgu Biiig is not the first quantitative poem in Turkish, for only a poetic tradition could have had this result, (Hence the fragments of poetry found in Mahmud i Kaygari’s Divan i Lugat it Turk ought to be subjected to a metrical analysis.) It should be noted, however, that the same words occur even more often without lengthened vowels. (There is an occurrence of dig with short / in § 332 above and also in the example quoted in this footnote ) Only a metrical analysis of the entire Kutadgu Bdig as well as SAiahut id Hakdvik. Divan i Hi kmc i etc. can answer the question it is generally assumed that the Old Turkish long vowels were still long in Karakhanidic Middle Turkish (Meedut Mansuroglu in Phihhgiae Turdcae Fundamcnta / Aquis Mattiacis 1959, p. 90), but if Karakhanidic had preserved the distinction, we should have expected this distinction to play a more important role in the poetry. If the rules which I have formulated are correct we must conclude that either the distinction had disappeared, though the poetic tradition still preserved some traces of it. or perhaps that the distinction between long and short vowels was at the poinl of disappearing, i.e. had an extremely weak phonemic status. fin his article Determination of Middie-Turkic long vowels through Sarud (Acta Orientalia Hungarica XX. 1967) Talat Tekin has tried to show that Karakhanidic long and short vowels may be distinguished by means of the motaqareb metre of Kutadgu Bdig. However, his results appear premature. One cannot deduce anylhing at all about long and short vowels by means of prosody, without first ascertaining the prosodical rules of the language in question. Talat Tekin says nothing about Karakhanidic prosody. He merely looked through “some 500 lines from the work" and found that "with the exception of one or two furkicised words, all the Arabic and Persian words are scanned properly". Now, the “one or two Turkiciscd words" arc rather numerous, and there is no reason why " I urkicising" a loanword should result in loss of vowel length, if the language had itself a distinction of short and long vowels. According to Talat Tekin the poet "never scans a short syllable long in Arabic and Persian words' . but as we have seen (§329 above) he does. 216 karakhanidic prosody §333 Apart from that twenty out of the twenty-seven Persian and Arabic words with long vowels mentioned by Tekin would scan tusi as well if the vowels had been short: din (thrice), iaj . ku\vi\ imtihan. uimdm, arguvan, zafydran, jadddd. kdd. and mtrfin because the syllable is already closed; ijaldyik. wrikat (twice), vafikin . and munafik according to §329 above; / aid , vafa, and kisru according to §327. “For the sake of further assurance [Talat Tekin] then checked the scansion of secondary long vowels'' and found that "all syllables having a secondary long vowel are scanned long ". He then quotes four instances of 17//, seven instances of kind and one instance of kordumi, kopgumi and kdrkumi. All these words end in two open syllables, and since the ntotaqdrcb metre does not allow two short syllables to follow one another, one of them must be scanned long. E.g. yiti is not possible in the metre, so it must be either or v, - , According to §327 above "any word boundary may optionally count as a consonant", so that we can read yiti# v - (Talat Tekin's yiti ). whereas there is no rule allowing us to scan the first syllabic as long. That is to say. in words having the syllabic structure found in the examples quoted by Tekin the last syllable must be scanned long and therefore such words tell us nothing about the vowel quality, If we look at words with secondary long vowels having a different syllabic structure it is no longer true that "all syllables having a secondary long vowel are measured long", E g, ilig "king" ( < iltig) may be scanned dig as well as dig. and dig is the commoner of the two (see the example quoted at the beginning of this footnote), I therefore cannot agree with Talat Tekin when he says, "there is no reason why we should not trust the poet in the scansion of Turkic words with primary long vowels". Talat Tekin concludes his article with three lists of primary long vowels, namely: "Vowel lengths of the first syllable", "Long vowels of non-first syllables", and "Vowel lengths in suffixes", However, "long vowels of the first syllable do not form a complete system " and are scanned "both long and short on different occasions". Moreover the poet “is riot very consistent in measuring non-first syllables of Turkic words", In other words Kutadgu Bdig cannot be relied upon as far as vowel length in "first" and "non-first" syllables is concerned. When it comes to "Vowel length >n suffixes", Kutadgu Bdig is apparently consistent and reliable, but unfortunately the great majority of the long vowels discovered by Tekin occur in positions where short vowels must be scanned long according to §329 above The first two examples are a^ima: and ayinu where i (belonging to a non-final open syllable following another open syllable in the same word) must scan long according to the said paragraph, The third example is afasit. Here again the second a must scan long according to §329. The it may be scanned long or short according to § 327. but since u - ^ is not possible in the motaqdrch metre il must be long here. The same two rules can be applied to all the following examples and we have to pass through no fewer than 131 of Tekin's examples before we come to suyurkdgu, where the long a cannot be explained by its position in the word. If we follow the scansion rules given in this chapter, we must scan either stiyurkagu u "*■ v v or suyurkagu# M — , which means that the word cannot be employed in the motaqdreh metre. So Ulug Hacib, when necessary, cut a heel and clipped a toe, as we say in Denmark, in order to include such words. — hesc observations should not be taken as a criticism of the general usefulness of prosodic research, but consistency over the whole metrical system is necessary in order to ensure a scientific result). XXVI THE ADAPTATION OF CLASSICAL PERSIAN PROSODY TO OTTOMAN TURKISH §334 The following description of Ottoman Turkish prosody 1 is an exposition of the general principles 2 . From the point of view of prosody, the most striking characteristic of Ottoman poetry is that no uniform rules can be formulated for scanning. It is written in a hybrid language consisting of a native West-Turkish and an adventitious Persian 3 clement, which at times is close to ousting the former. The Persian element follows the rules of Classical Persian prosody completely. The native element follows other rules which are similar to, but not identical with the rules given for Karakhanidic in the previous chapter 4 . §335 Now consider the following lines by jT Baki 1 A recent work dealing with Ottoman Turkish prosody is Walter G. Andrews: Ottoman Poetry (Minneapolis and Chicago 1976), The chapter on prosody is an original and important contribution to Oriental scholarship Cf. my forthcoming review in Ana Orwntafia. 2 If a second edition of this work should be published, it will contain a complete description of Ottoman prosody. 3 The Persian element again can be said to consist of a native Persian and an adventitious Arabic clement, but for our purpose this distinction is unnecessary. 4 In this respect Ottoman Turkish prosody is radically different from Urdu prosody. Urdu, too. is a hybrid language, but metrically the same rules apply to the native and foreign elements. 218 OTTOMAN PROSODY §335 ey pay'* bend i dam'- geh i: kayd t ndm u neng id kev he- vd yi me$ga- le i: dehr i bidireng mm f ** a:n ol gu nu: ki ahi- r olup nevba- hdr i Somr berg i: ha- zdna donse gerek ru yi lalereng dhir me- kamh olsa gerek ciir^a gi:bi hdk devrdn e- linden erse gerek cam i ‘iaysa seng insdn o- dur ki dyi- neves kalhi saf ola sinende n eyler 5 6 dde- ar rn iseh kine i: pel eng*' “ “ ivy — w — V-/ v* O' — w — “O (thou who art) fettered in the world's bondage of fame and glory! How long (this) lust for dealings with fleeting time (i.e. life in this transitory world)? Remember that day when life's springtime is over and the tulip- coloured cheeks (lit. face) must turn into autumn’s leaf. Your last abode like (that of) the dregs 7 must be the earth. A stone from fate’s hand must hit (and break) pleasure's cup. A man is he, whose heart is clear like a mirror. What is panther’s passion doing in thy breast if thou art a man”? (Rhythm in §§237-238.) The First beyt could have been taken from a Persian poem. A Persian would read it: ei pay - band e dam '- gah e qeid e nam o nang id kei ha- va ye masqa- le ve dahr e biderang in the remainder of the passage Turkish elements occur, but Persian elements are more frequent. Everywhere, the Persian element can be scanned according to the rules of Persian prosody. §336 Note that the Persian ezdfe (Turkish izdfet) and the Persian conjunction o (see Chapter VII) are in Turkish subject to vowel harmony: ezdfe is realized as i and / (and in Modern Turkish also ii and u). Persian o “and” 5 For ne cyler. The reduction of ne to tt is not restricted to poetry. On the contrary, though not allowed in standard prose it is very common in the spoken language: napabm'l nolacak? etc. See G.L. Lewis: Turkish Grammar (Oxford 1975) p. 74. Cf. also §§79-81 above. 6 My transcription is based on Modern Turkish orthography and follows as closely as possible Ferit Dcvellioglu; Osmaniica Tiirkye Ansikbpedik Lugal (Ankara 1962). 1 distinguish between j « and kS h. Metrically shortened long vowels are shown with * (d, w). Metrically lengthened short vowels are shown with (/;, u;, a:}. 7 The dregs are namely thrown on the ground. §338 imale 219 becomes ii and u. After long vowels they arc both pronounced with glide consonants and without vowel harmony, yi and vw respectively, (Cf. §54 footnote 6, and §55). §337 The passage quoted in §335 contains the following native Turkish clement: a:h ol gii- nii : olup -a douse gerek -in olsa gerek gi'bi e- linden erse gerek -a 0- dur -t oia -hde n evler -p isen “ ” w V ^ V ^ v v — Here, as in Karakhanidic, the rhythmical effect is brought about by the opposition of open and dosed syllables, e.g. o-fup ^ ge-rek e-lin- de-ner-se w- There are three cases where open syllables must be scanned long: a:n ol (or an of) - (instead of an of v -). giinu: and gi'bi. The three short vowels a , ii and i have been metrically lengthened to a:, ii: and i:. The metrical lengthening of a vowel is termed imctle "inclination (towards length)" 8 9 . §338 As a rule, the Persian element occupies the major part of an Ottoman poem. In the following example less than one third is Turkish: hikmet ge- endbfe dii$mez di- te$vi$ i r H o mim i:se ii S'akf i: be - l ii tab V ehfi dil ii: tab ft ki ahkam t felekten $eri bdha - her evler bu endive- ye zird bu da: hi: be- ter 10 evler *r 8 The term emote in Persian has a quite different meaning. See §§145 and 146, 9 Cf. §120. 10 Cf. §111. 220 OTTOMAN PROSODY §338 Turkish: o i:se -ten e vler 9‘ du$mez m */ hu -ye hu da: hi: vyier - — “ w O' w — ■ — w u — — “If philosophy be that science which makes human mind and reason acquainted with the decrees of Heaven (i.e. future events), the wise (lit. possessor of heart and character) will not take to such speculation (lit. this thought). For this (lore) makes the confusion of heart and mind more confounded still". ^ NefSI (Rhythm in §§220-222.) Here we have three instances of imdle ( i:se , da; hi:). §339 Of course, it is possible to find poems with a larger native element. More than half is Turkish in the following lines from the ode in praise of the boys of Adrianople by ^ r Mesihi. Q ur Turkish : birt:si:dir halt! o: d- *5 w ■1^ * gohiiiler V> dinir ' ' her ge:- ce mihmdn ne sihr etmis »+ m-m. § ** M vuzunde: zul- ■r f i serke$ ki a:ha: sah- n i gulden ol- du ]l d te$ hiri:si:dir gonutler e:- o: dinir ge> ce ne eimi$ a ha: viizunde: w oh du w ~ \J — *“ “ \_t ^ II I have scanned with hanxze (Turkish hvmztr) according to §45, Jl is also possible, but {to my ear) less naiural to read gonulie:r e dtmr with imdle + but since gulden is Persian 11 is not possible to read *gii/$e:n with imdle §340 1MALE 221 "(And another) one of them is Halil, that calamity of the soul l2 . Of hearts he makes guests every night. What magic have his rebellious locks wrought upon his face, so that for him the lire has become a bed of roses 1-1 "! (Rhythm in §§194-196.) i Here we have no fewer than eight instances of imdie : More than half of the open syllables in the Turkish portion are subject to imdie, and exactly half of the metrically long syllables in the Turkish portion are made long by imdie. In native Turkish words, theoretically, all vowels in open syllables may be metrically lengthened, but in practice it is hardly possible to have a higher proportion of imdie than in the above example. (In a line with only Turkish words and no closed syllables, there would be nothing to show the rhythm.) In spite of the high proportion of imdie , we cannot consider lines such as the above a relapse into syllabic poetry. If that had been the case, we should have expected to find now and then closed syllables where the metre requires short syllables, and that is never the case. §340 There is not and probably never was anything in the pronunciation of Anatolian Turkish to justify imdie. It can only be considered an emergency measure, and for once it is appropriate to talk of i dalles productae metris causa. So,, naturally, one of the hallmarks of "fluency” — ravdni 14 the Persians 12 l.e. who by his beauty destroys one's peace of mind This expression was so common lhal in Modern Turkish diet alone has come to mean ‘'person of bewitching beauty" (gurc/ligf He inrnni fa f kina fwiren Turkic Sozliik). !3 The boy's name immediately makes the reader think of Abraham, called Halil iir Rahman "Familiar of (God) the Merciful" (j^ll J-U-) Abraham is said to have been so hospitable that he would not spend a night without having enleriained some guesl, just as the boy makes hearts his guests every night. When king Nimrod wanted to have Abraham thrown on the fire, it was miraculously changed into a bed of roses. For little Halil, too. the fire has been changed into a bed of roses. This t take to mean that the boy, who must have been rather naughty, used to cause anger, but by his charming looks and locks changed the fire of wrath into roses of favour. Gibb (It p 253) has a neat translation of this verse: "And one Khalil, that torment of the spright,/ Ht taketh hearts to be his guests each night, Upon his face what spell has cast his hair., That thereunto the fire is garden fair'' * 14 See §50. In Ottoman Turkish revan does occur in the sense of "fluent", hut it is not common, and Ottoman and Modern Turkish revani (from Persian ^sj rouqani) is a kind of sweet, (a light cake, made of tine wheat grit (trmik), eggs, etc. and soaked in sugar water). So ] prefer not to use the terms revan and revani here. 222 OTTOMAN PROSODY $340 would call it— is the relative scarcity of imdle, and exactly this scarcity or even absence of imdle is a most conspicuous quality of great formal masters like Baki 15 and fX Nedim. I * §341 After Baki (d. 1600) the Ottoman poets tried to resort as little as possible to imdle. The easiest way of escaping imdle was to use as few native Turkish words as possible, and much “Turkish” poetry has been written which is really Persian except for some Turkish case endings and an occasional form of "to be” or “to do”. But the real master craftsman of Ottoman Turkish poetry was he who could use native Turkish words freely and keep imdle under control. In the following lines by Ncdini more than half is 7 urkish and all the same there is not a single instance of imdle. jism i pdkin dediier hem ter imis hem berrdk depeden tir- naga dek giil gibi biliu- r' gibi nigehin boy- le neden has- iadir ey jai- b ’ sen in gdzlerin be:- m i ezefden beri mahnnt- r’ gibi Turkish: -in dediier imis depeden in 4 - naga dek gibi . gibi t ^ f ** - m hoy- le neden - dir sen in gorier in w -den beri gibi w w w — — WO ' w w ” ' — T “Your flawless (lit. pure) body, they say, is both fresh and sparkling. (Yea,) from top to toe (lit. nail) (it is fresh) like a rose (and sparkling) like crystal. O (my) mischievous (friend), why are your eyes so languorous? It seems your eyes have been drunk since the Feast of Eternity l6 ’\ (Rhythm in §§211-212.) 15 In the lines by Baki quoted in §335 there is less than one imdle to the bevt. 16 The Feast of Eternity is a mystical image. To the discerning eye of the mystic, the world is God's eternal Feast of Love. In a letter Dr* A. Schimmel suggests that hezm i ezel alludes to yevm i elest (C— h ^ : the beginning of creation, ef, qor'art VII, 172). §343 TURKICISMS 223 §342 In the example above the native Turkish element amounts to 33 out of 60 possible syllables, but there arc several of the Persian words, loo, which have been completely naturalized in Turkish, thus hem, gift , and hast a. When separating the native Turkish clement in Ottoman poetry from the Persian one, it must always be borne in mind, that a varying portion of the Persian element was not at all felt by the Turks to be foreign. We have no certain criteria for deciding exactly when a loanword is completely naturalized, an one would always have the doubt whether a word which is naturalized now was already naturalized w hen the poem in question was written. A good reason for not trying to include naturalized words in the Turkish clement, is the fact that, in poetry, even completely naturalized words were still pronounced and scanned according to foreign rules. §343 One should expect the most thoroughly acclimatized and most commonly used Persian words in Turkish to be occasionally pronounced and scanned like native Turkish words, and it does indeed happen l7 . Fastidious poets like Baki and Nedim carefully eschew treating even the commonest Persian loanword as Turkish, but other poets and not the least interesting ones do occasionally ignore the foreign origin of such words. In the following two examples by «« Galib Dede the words ^Jjand must be scanned according of their ordinary Turkish pronunciation, vakil and madem instead of the proper Persian 18 vakt and madam. mehd t\ re kalmaz bu Mr hat /- uyu: ki ey revise car- te gerdis et- semenber h i cenber mez ahter 17 Cf. what was said about Persian loanwords in Urdu poetry in $§279-284 (especially 284) and 31 1, 18 I.e. according to the traditional Turkish pronunciation of Persian poetry. Incidentally both the "Persian" words arc Arabic, 1 call them Persian according to §334 footnote 3. 224 OTTOMAN PROSODY §343 seyretse- sevf i: ga he a: vakit- ma as n o- te n evler b olursun “Sleep in (your) cradle, for, O darling, the wheel of fate will not remain in this wise. The stars will not (always) turn in one way. Just look (and see) what they will do in a short time. You will become a millstone for the torrent of grief". (Rhythm in §§252-253.) I >. < ^ I . . i ( ^il . „ . . £1/^7 & \r * 1 JjYJ a U bunda modem rmh w mihr u: Sdsikdn ey- as'/nan ey ler sema 5' ler semd V “While the Lovers (i.e. Whirling Dervishes) here are doing (their) circular dance, moon and sun and sky do (their) circular dance". (Rhythm in §§201-203.) § 344 In §332 we saw that overlong syllables are not distinguished in Karakhanidic Turkish prosody. In Ottoman Turkish prosody on the contrary overlong syllables are prosodically significant both in the Persian and in the native Turkish portions 19 . However, I suspect that this metrical significance of the overlong syllables was artificial. One does now and them come across metrically disregarded overlong syllables. if / m r # ' * * 0 9 _ M m # m I l» vapip dunyd rt 2 0 ne jtr>av fT - v yap ■ ba/tk boston ne kddir kirn- evtn vi ra- ti ne: sedddd 1 ' kokar derier se bir nokta: n edersi r ha- bind etti: fesadm ba:- dive: 22 hdzd ne i: dini bu sek 1 2 1 billdh Si ma Sfumdur 2 1 kitdbulldh w w 19 Cf. Chapter V above. 20 Against §88. 21 Againsl §36. (In Modern Turkish original long vowels in loanwords are often retained in open syllables, but shortened in dosed syllables. Cf. G.L. Lewis, op. eit. p. 14.) 22 Optative. §344 EXERCISE 225 “Building up the house of the world, you lay waste the house of the faith. By God. in this way neither Pharaoh did make (his pyramid) nor §eddad construct (his palace and garden of Irem :3 ). The fish (begins to) stink at the head (i.e. corruption starts at the top)', they say. The ‘head' of (our) corruption is (well) known. Not (as far as) one iota can one say: This (state of affairs agrees with) the Book of God' Uveysi (Rhythm in §§191-193.) EXERCISE : Before taking leave of those readers who have had the patience to follow me thus far, 1 want to repeat and expand the advice given in §172: Henceforth never read a line of Oriental poetry without ascertaining its rhythm. A correct knowledge of prosody is an important tool for deciding the correct reading and interpretation of texts written in the defective Arabic alphabet, a tool which no earnest student of Oriental lore can afford to ignore. As an exercise try to scan the following lines by JIjJL- Suleyman Dede: “God’s name let us mention first. Incumbent is that in every work upon every slave (of God. i.e. created being). God's name: whoever mentions that first, his every work will God make light for him. God's name let be the start of every work so that vain may never be its end. With every breath God's name say complete. (Then) by God’s name shall (your) every work become complete''. ## * 23 The Persian forms are saddad and eram. APPENDIX ONE REFERENCE LIST OF METRES The following is a complete list of all Classical Persian metres (and permitted variations) that I know of 1 , arranged in the quasi-alphabctical order described in § 189. The basic rhythms are numbered and the permitted variations unnumbered. The robust rhythms are marked off with 'R, cf. §255. After each rhythm its traditional designation has been given with the terms indicating overlong finals added in brackets 2 3 * , cf. §42 (but see remark on no. 7 below). This is followed by references to the relevant paragraphs in Part Two and/or additional references to literature and to Elwell-Sutton. Finally Elwell-Sutton's code-number has been given (bracketed when his classification seems artificial) cf. §§165, 186 and 245. In case of permitted variations the reader is referred to the serial number of the corresponding basic rhythm 5 . Doubling is only indicated when the form of a particular permitted variation depends upon it. In Appendix One the following abbreviations are used: al. aslam (e) 1)1 mb. mahzuf mx. ma.xbun (e) as. a slam (e) 1^1 mn. mosamman (e) p.v. permitted variation (of) b. bahr ( e ) mq. maqsur r. ramal (e) h. hazaj (e) ms. mosaddas ( e ) s . sarrf (e) m. mojtass (e) mt. matvi (ye) X. xqftf (e) ES stands for Elwell-Sutton. References to literature as explained in § 175. The traditional designations given are everywhere those which would be used to describe a single mesra In case of complete rhythmical identity of the two halves, a beit is named exactly as a single mesra *i except that 1 The irregular variations of bahr e raja: found in M 262 are not included in the list, nor is M 2121 which is realty a syllabic rhythm. (In both cases we have to do with Greek molamma^di). 2 The sign / before the term indicates that it should be substituted for the preceding term. 3 Permitted varialions are not given for rhythms not known to occur outside the handbooks of prosody. 228 APPENDIX ONE l — \J XJ w w ^ w \y x+t xy — xy xj the terms sadr, hasv and zarh. if' occurring, must be changed to sad re in. has a vein and zarbein respectively. If there is not complete identity, i.e. if various permitted variations occur, the designation of the complete be it is obtained by combining the terms ebtedd \ sadr , baruz, zarh, va, on etc. with the terms used to designate each single mesrab. Cf. $$42. 177, 185, 223 and 242. 1 motaddrek e mn . nix. (mozal Ji-U ). § 1 72. M 2254. (3. 1.3(4)). ^ ^ — b. motaddrek e mn. nix maqtu V oz-zarb (el-mozdl ) , p.v. I . w- ^ w - — w- b. motaddrek e mn. mx. os-sadr va mx. va maqtu b ol-hasv va mx. oz-zarb (el-mozdl), p.v. 1. v ^ ^ ~ — — b. motaddrek e mn. mx. os-sadr va mx. va maqtu b ol-hasv va maqtu b oz-zarb (el-mozdl), p.v. 1. 2 w-v- w-v- b. kdmel e morabbab ,*jy* e salem (e mosbaq). M 238. (5. 3.5(2)). 3 b. kdmel e mn. sdlem (e mosbaq). §§261-262. (5. 3,5(4)). 4 w w - w - w — v - b. r. mn. maskul va salem e mh. (Imq.). ES 111. 5.3.15. 5 w - w -v — — b. r. mn. maskul va salem ( e mosbaq ) . §§248-249. 5.3.16. 6 - v/ — — b. r. mn. maskul os-sadr va salem va maskul ol-hasv va as. oz-zarb (el mosbaq). ES 111. 5.3.14. w ^ - w - ^ w - v - b. r. mn. maskul os-sadr va salem va mosabbas e makfuf ol-hasv va mh. oz-zarb (Imq. oz-zarb), p.v. 4. wv/-w -V — — w - v — b. r. mn. maskul os-sadr va salem va mosabbas e makfuf ol-hasv va salem oz-zarb (I mosbaq oz-zarb), p.v. 5. ^ w - w - ^ ^ — b. r. mn. maskul os-sadr va salem va mosabbas V V V V Xy XJ — \y — Xy VV — KJ \J xy ~ uu w W W — KJ \J W \y xy \J — xy \J xy xy xy — xy Xy xy — xy e makfuf ol-hasv va as. oz-zarb (el-mosbaq), p.v. 6. 7 uw-u — ^ w - w — b. kdmel e morabbab e moraffal. M. 963, 3111- 3113. (5. 3.6(2)). There are no permitted variations of this metre. The term moraffal indicates the addition of a final long syllable to the basic foot. The corresponding overlong foot is termed motavval Jjlx*. It is extremely rare and has therefore in most cases not been given. 18 REFERENCE LIST OF METRES 229 8 V V \J \J I ~~ V — WV — — WW W V/ w V 1 10 I I 12 13 V ^ w W W u >J v* V V ~ V V 14 vy w 15 w \J \j 18 'w' vy v/ v vy vy \y vy w vy V w \y vy vy vy Irregular variation of 2. (Against § 152.) - h. madid e mn. mx. ( mozdl ). ES 97. 3.1.07 (2). - ^ ^ — — b. madid e mn. mx. maqtuS oz-zarb (el- mozdl), p.v. 8. - ^ w - w - w h. r. ms. mx. os-sadr va maskul ol-haSv va salem oz-zarb (el-mosbaq). ES 111. (5.3.03/09.) — b. r. morabba ? e mx. (mosbaq). ES 96. 3.1.08. — b. r. ms. mx. mb. ( jmq .). §§213-214. 3.1.1 1. — b. r. ms. mx. (mosbaq). ES 96. 3.1.12. - ^ ^ — w ^ — w ^ - b. r. mn. mx. mb. (jmq.). §§211-212. 3.1.15. - w — w — ^ ^ — b. r. mn. mx. ( mosbaq). § 2 10. 3.1.16. In M 94 this rhythm occurs as an irregular variation of 54. b. r. mn. mx. marbu S (e mosbaq). ES 97. 3.1.14. ■- ^ w — ^ ^ — - b. r. mn. mx. mqjbuf GI 185, ES 97. 3.1.13. - b. r. mn. mx. as. (e mosbaq), p.v. 13. — b. r. mn. mx. mosaWas oz-zarb (el- mosbaq), p.v. 14. - w ^ — w - w - b. jadid e ms. mx. (mozal), §244. ES 104-105. (4.5.04/08). •- ^ v — w- b. jadid e mn. mx. mb. (jmq.). ES 105. (4.5.04/11). — b. jadid e mn. mx. as. (e mosbaq), p.v. 18. v~- b. madid e mn. mx. os-sadr va mx. va salem ol-hasv va mx. oz-zarb (el-mozdl), p.v. 8. - uu — w-- In M 94 this rhythm occurs as an irregular variation of 54. b. madid e mn. mx. os-sadr va mx. va salem ol-hasv va maqiu V oz-zarb ( el-mozdl) , p.v, 8. — b. r. mn. mx. os-sadr va mx. va salem ol-hasv va mx. oz-zarb (el-mosbaq), p.v. 14. b. r. mn. mx. os-sadr va mx. va salem ol-hasv va mosaWas oz-zarb (el-mosbaq), p.v. 14. — b. r. ms. mx. as. (e mosbaq), p.v. 1 1. vy \y \j vy w vy ■vy vy vy w w vy — vy vy \y vy vy v v v vy 230 APPENDIX ONE 19 19 20 21 22 23 vy vy vy w w vy vy V v v sy vy vy vy vy vy \j vy vy vy V vy vy vy vy vy v/ vy vy w ^ vy vy vy vy v vy vy vy v/ vy v/ vy vy vy v/ w vy vy ^ vy — w ^ 6. r, ms. mx. mosaWas oz-zarb ( el-mosbaq ) , p.v. 12. — w w — w ^ - b. r. mn. mx. os-sadr va mx. va mosa *1*1 as ol-hasv va mx. oz-zarb ei-mh. (/el mq.), p.v. 13. -- w — w — b. r. mn. mx. os-sadr va-z-zarb va mx. va mosa $ Sas ol-hasv, (b. r. mn. mx. os-sadr va mx. va mosaWas ol-hasv va mx. oz-zarb el-mosbaq), p.v. 14. — wv — b. r. mn. mx. os-sadr va mx. va mosa ? ^as ol-hasv va majhuf oz-zarb (el mosbaq ) , p.v. 16. — ^ ^ — — h. r. mn. mx. os-sadr va mx. va mosa Was ol-hasv i’j as. oz-zarb ( el-mosbaq) , p.v, 13, — v ^ — b. r. mn. mx. os-sadr va mx. va mosa^^as ol-hasv va mosa Was oz-zarb ( el-mosbaq ) y p.v. 14. — w- v - b. x. morabbaS e mx. (mozal). ES 104. 4.5.08. — v,v/- b. x. ms. mx. mh. (jmq.). §§230-232. 4.5.11. -- ^ - w- w ^ — h. . v. ms. mx. (mozal ) . ES 104. 4.5.12. — ^ w V — w-v,- b. x. mn. mx. (mozal). M 243-244, 965, 1176, 2098, 2255-2259, 3139-3140, 3203; ES 105. 4,5.16. — v ^ - - b. x. ms. mx. majhuf (e mosbaq). ES 104. 4.5.09. — — b.x. ms. mx. as. (mosbaq), p.v. 20. — w - w b. x. ms. mx. mosa Was oz-zarb (el mosbaq), p.v. 21, — v w - 6. x. mn. mx. os-sadr va mx. va mosaWas ol-hasv va mx. oz-zarb (el-mozdl), p.v. 22. — — ^ v - b. motaddrek e mn. mx. os-sadr va maqtu ? va mx. ol-hasv va mx. oz-zarb (el-mozdl), p.v. I. — — wv- — b. motaddrek e mn. mx. os-sadr va maqtu $ va mx. ol-hasv va maqti'W oz-zarb (el mozal), p.v. 1. — — w w — v/^-6. madid e mn. mx. os-sadr va maqtu V va mx. ol-hasv va mx. oz-zarb (el-mozdl), p.v. 8. — — ^ w — — b. madid e mn. mx. os-sadr va maqtu S va mx. ol-hasv va maqtu^ oz-zarb (el-mozdl), p.v. 8. — ^ w - h.r. ms. mx. os-sadr va mosa V Ws ol-hasv va mh. vy “ vy vy — vy vy " v/ vy vy vy vy vy vy vy ■" vy vy — v> v/ vy vy vy vy vy vy vy vy vy vy vy (jmq.) oz-zarb el-mx., p.v. 11. Or ^ w -) b. motaddrek e mn. mx. os-sadr va-z-zarb (el-mozdl) va maqtu V ol- hasv, p.v. 1. — w w — fa, r. ms. mx. os-sadr va-z-zarb va mosa $ Ws 23 REFERENCE LIST OF METRES 231 v v \j w w W v c/ v/ c/ vj U M w w ol-hasv. (b. r. ms. mx. os-sadr va mosaS^as ol-hasv va mx. oz-zarb el-mozdl ), p.v. 12. — ws*- b. r. mn. nix . os-sadr va mosa^^as va mx. ol-hasv va mh. (fmq.) oz-zarb , p.v. 13. -- b.r. mn. mx. os-sadr va-z-zarb va mosa Was va mx. ol-hasv, (b. r. mn. mx. os-sadr va mosa Was va mx. ol-hasv va mx. oz-zarb el mozal ) . p.v. 14. ^ w — - b. r. mn. mx. os-sadr va mosa^^as va mx. ol-hasv va majhuf oz-zarb (ef-mosbaq), p.v. 16. — — h. r. mn. mx. os-sadr va mosa^Ws va mx. ol-hasv va as. oz-zarb ( el-mo sbaq ), pv. 13. — b. r. mn. mx. os-sadr va mosa *t Sas va mx. ol-hasv va moiaWas oz-zarb (el-mosbaq), p.v. 14. ^ ~ b. jadid e ms. mx. os-sadr va-z-zarb va masalas ol-hasv, ( b . jadid e ms. mx. os-sadr va mosaWas ol-hasv va mx. oz-zarh el-mozdl), p.v. 17. w - ^ - w w - b. jadid e mn. mx. os-sadr va-z-zarb el-nth . ( jel-mq.) va mosa Was va mx. ol-hasv, p.v. 18. - - ^ v - — h. jadid e mn. mx. os-sadr va mosa Was va (jel-mq.) va mo s abb as va mx. ol-hasv, p.v. 18. - - — f - v — ^ ^ - b. madid e mn. mx. os-sadr va maqtu $ va sdlem ol-hasv va mx. oz-zarb (el-mozdl), p.v. 8. - - — j - ^ — — b. madid e mn. mx. os-sadr va maqtu V va sdlem ol-hasv va maqtu $ oz-zarb ( el-mozdl) . p.v. 8. — h. r. mn. mx. os-sadr va-z-zarb va mosa V . mozareS e ms. makfuf e majbith ( jab tarn) . ES 105. 4.6.10. v -v/ - v w — b. mozare S e ms. makfuf e mb. ( mq .). ES 105. 4.6.11. - v - w v — w - v - b. mozare 5 e mn. makfuf e mb. (fmq.). ES 106. 4.6.15. - b . mozare 5 e ms. makfuf e abtar ( fazall ). ES 105. 4.6.09. v - v- v — ^ -v— b. mozare ¥ e mn. makfuf va mb. (fmq.). ES 106. 4.6.07(2). w - w — v - b. mozare 5 e ms. makfuf va salem va majbub (fahiam). ES 90, 116. (9.4.10). • w v — w -v — b. mozare b e mn. makfuf va salem ( fmosbaq) . ES 106. (9.4.16). ^ w - - b. mozare S e ms. makfuf va salem va abtar (fazall). p.v. 61. ^ ^ — w - ^ - b. mozare 5 e mn. mak fuf va salem va axrab va mb. (fmq.). p.v. 63. w — b. motaqareb e ms. mb. (/mq.). ES 89. 1.1.08. w - b. motaqareb e mn. salem os-sadr va salem va maqbuz ol-hasv va abtar ( jazall ) oz-zarb. ES 115 (9.1.10). w — ^ - w - - b. motaqareb e mn. salem os-sadr va salem va maqbuz ol-hasv va al. oz-zarb (el-mosbaq). ES 3 15. (9.1.11). v v ^ — w - b. moqtazab e mn. mx. va salem (fmosbaq). ES 90. (1.1.08(2)). - b. motaqareb e ms. salem (e mosbaq). ES 89. 1.1.09. v ^ v- b. motaqareb e mn. mb. (fmq.). §§178-183. 1.1.11. v t), motaqareb e mn. salem (e mosbaq). §§ 173- 174, 177. 1.1.12. - w — b. motaqareb e mn. abtar (fazall). ES 90. 1.1.10. w w w b. tavil e mn. salem (e mosbaq). ES 90. (1.1 .07(2)). 236 APPF.ND1X ONF 77 V — V/ 77 78 79 80 81 82 83 84 85 86 87 "" ^ v/ v/ vj vj w 88 — b. moqtazab e mn. mx. os-sadr va maqtu V e morqffal va mx. ol-hasv va ml. oz-zarb el-moraffal , p.v. 59. — w b. moqtazab e mn. mx. va maqtu*) e moraffal , p.v. 59, — 6. h. morabba V t j m/i, (/mq.J. ES 92. 2.1.07. / w — />./;. m«. .?£/*?» va m/i. (jmq.). ES 92, (2.1.07(2)). - w _ - _ h. b. morabba ? ^ sdlem (e moshaq ) . ES 92. 2.1.08. - ^ ^ — b. h. ms. mb. (jmq.). §§194-196. 2.1.11. — v b. h. ms. sdlem ( e moshaq) . ES 92. 2. 1.12. — v b. b, mn. majbuh (jahtam). ES 92. 2.1.14, u v w — b. h. mn. mb. (jmq.). F.S 92-93. 2.1.15. u w w b. b. mn. sdlem (e mosbaq). §§191- 193. 2.1.16. w — b. mozdre 5’ e morabba $ e sdlem (j mosbaq). ES 116. (11.1.08). w — w b. mozdre ? e ms. sdlem (l mosbaq). ES 117. (11.1.12). — w -w — b. mozdre *> e mn. sdlem (/mosbaq). ES 117. (11.1.16). — w w — w w — b. b. mn. sdlem va axrab va makjuf va mh. (jmq.), p.v. 54. — v v — b. b. mn. sdlem os-sadr va axrab va sdlem ol-hasv va ahazz oz-zarb ( el-mosbaq ) , p.v. 54. — w - w - b. qarib e ms. e sdlem va axrab va mh. (jmq.), p.v. 55. — „ w w ^ — b. h. mn. sdlem va ahazz va makjuf va mb. (mq.), p.v. 57. — v — b. h. mn. sdlem va ahazz (e mosbaq), p.v. 57. — w v - - b. h. mn. sdlem va moxannaq va axrab va mb. (jmq.), p.v. 54. — b. b. mn. sdlem os-sadr va moxannaq ol-hasv va ahazz oz-zarb (el-mosbaq), p.v. 54. ^ ~ v-w- b. rajaz e ms. ml. os-sadr va mx. ol-hasv va-z-zarb (el-mozdl), ES 110. (5.2.08/04). w ^ - w - w - ^ ^ - b. rajaz e mn. ml. os-sadr va-z-zarb ( el-mozdl ) va mx. ol-hasv. ES 111. (5.2.08/08). 89 106 REFERENCE LIST OF METRES 237 90 -uw- o-o- -vv- b. raja: e ms. mi. as-sadr va-z-zarb (el-mozal) va mx. ol-hasv , ES 110. 5.2.12. 91 ~ o o - w-w- -wo- w - o - h. raja: e mn. ml. va mx. ( mozdl ) . §§250-251. 5.2.16. 92 - o o - o - w - - o o - - o o - b. raja z e mn. mi. os-sadr va-z-zarb (el-mozdl) va mx. va ml. ol-hasv. ES 111. (5.2.12/04). 93 -oo- o-o- — w- b. s. ms. ml. va mx. va ml. maksuf ( jmouquf). ES 1 10. 5.2.11. -oo- o-o- v-w- b. rajaz e mn. ml. os-sadr va mx. va maqtu ? ol-hasv va mx. oz-zarb ( el-mozal ), p.v. 91. 94 - o o - w — b. monsareh e mo r abba ^ e ml. os-sadr va mx. oz-zarb el-maksuf ( el-mouquf) . ES 110. 5.2.07 95 -ow--ww-6 . rajaz e mo r abba ^ e mi. (mozdl). ES 99-100. 3.4.8. 96 -wo- -ww- w-w- - o w - b. rajaz e mn. mi. os-sadr va-z-zarb (el-mozal) va ml. va mx. ol-hasv. ES 111. (5.2.04/12). -wo- - o o - w - o b. raja: e mn. ml. os-sadr va ml. va mx. ol-hasv va maqtu V oz-zarb (el-mozal). p.v. 96, 97 - o o - - wo — b. rajaz e morabba 5' e ml. morajjal. ES 100. 3.4.9. 98 -ww- - o o - -oo- b. rajaz e ms. ml. (mozdl). ES 100. 3.4.12. 99 -wo- -wo— -oo — b. rajaz e ms. ml. morafjal. M 53, 1 022, ES 100. 3.4,13. 100 -oo- -oo- -oo- -oo- b. rajaz e mn. ml. (mozdl). §215. 3.4.16. 101 -oo- -oo- -oo- — h. rajaz e mn. ml. ahazz (e mosbaq). ES 100. 3.4.14. (The example quoted by Elwell-Sutton consists of eleven long syllables. It could therefore equally well be considered a variation of 167). -oo- -oo- -oo- b. rajaz e mn. ml. os-sadr va-l-hasv va maqtu*) oz-zarb (el-mozal). p.v. 100. 102 -oo- -oo- -o- b. s. ms. mt. maksuf (jmouquf). §§216219. 3.4.11. 103 -oo- -oo- - 0 - 0 - 6 . s. ms. ml. morafjal ( motavval ) . GUI 344, ES 103-104. (4.4.05/08). 104 -oo- -oo- -o — b. badil e ms. mt. (mosbaq). ES 103. 186. (4.4.05/07). 105 -oo- -oo- — b. s. ms. mt. as. (mosbaq). ES 100. 3.4.10. 106 -oo- -oo — -oo- -oo — b. rajaz e mn. mt. va mt. morajjal. M 3196. 3.4.09(2). 238 APPENDIX ONE 107 V ^ V w \y ^ .- - ^ w — - v v b. rajaz e mn. mt. os-sadr va ml. moraffai va ml. ol-hasv va maqtu S oz-zarb el-moraffal , p.v. 106, . - - v/ w b, rajaz e ms. ml. os-sadr va-l-hasv va maqtu 1 ! oz-zarb ( el-mozd ! ) , p.v. 98. , - -vv- b. rajaz e ms. ml. os-sadr va-l-hasv va maqtu*! oz-zarb et moraffai , p.v. 99. - - - w v w - b. rajaz e mn. mt. os-sadr va-z-zarb ( el- mozal) va ml. va maqtu*! ol-hasv, p.v. 100. . - - ^ v — w w — b. rajaz e mn. mt. os-sadr va mt. moraffai va maqtu ^ ol-hasv va mt. oz-zarb el-moraffal , p.v. 106. / - - w v> b. rajaz e mn. mt. os-sadr va mt. va maqtu S' ol-hasv va maqtu *! oz-zarb ( el-moza! ) , p.v. 100. ■ - - — b. rajaz e mn. mt. os-sadr va mt. moraffai va maqtu V ol-hasv va maqtu S oz-zarb el-moraffal. p.v. 106. 107 ^ “ v — $7 . basit e 108 ~ w ” — v 4.4.12. 109 — ~~ — O' ^ V_> “ “ §§233-234. 4.4.13. no — ^ ~~ w w — - b. monsareh e mn. mt. maksuf "f fmouquf ) . ES 104. 4.4,15. -^w- - w - - v - b. monsareh e mn. mt. va mt. maksuf (mouqujj va mx. va mt. maksuf ( fmouquf j (or b. basit e mn. mt. va salem (f mozal) va mx. va salem (/mozal)), p.v. 113. Ill - v ^ - -w-v — b. monsareh e ms. mt. ahazz (e mosbaq). ES 103. 4.4.10. - v w — -w-w b. monsareh e ms. mt, os-sadr va-l-hasv va maqtu *1 oz-zarb (el-mozdl), p.v. 108. - vu- - w - v -b. monsareh e mn. mt. os sadr va mt. va maqtu *! ol-hasv va manhur ( jmajdu *!) oz-zarb, p.v. 109. 112 -v/u- - ^ — h. m. e morahbaS e mt. (mozal). ES 103. 4.4.08. 113 -vw- - v - -wv- -v- h. monsareh e mn. mt. va mt. maksuf (fmouquf) (or b. basil e mn. mt. va salem (j mozal)). §§235-236. 4.4.07(2). -v>^- - v, -v-b, monsareh e mn. mt, os-sadr va mt. maksuf (fmouquf) va maqtu*/ ol-hasv va maksuf (fmouquf) oz-zarb (or b. basil e mn. mt. va salem (/mozal) va maqtu*! va salem (/mozal)), p.v. 1 13. 1(7 REFERENCE LIST OK METRES 239 i 14 O \J 1 15 KJ \J O' w o' w 116 I 17 w o' W O' - b. rajaz e morubba *} e moraffal e mt. M 1280, 1362-1363, 2071. 2246, 3033-3036. ES 100. 3.4.5(02). - - w w — — h. raja: e mn . moraffal e mt. M 1269, 3007. 3.4.05(4). - - ^ — b. raja z e mn. moraffal e mt. os-sadr va-l-hasv va maqtu *i oz-zarb , p.v. 115. ■ - ^ ^ — b. rajaz e mn. moraffal e mt. os- sadr va-z-zarb va mt. va maqtu*} ol-hasv , p.v. 1 15. b, rajaz e mn. moraffal e mt. os- sadr va mt. va maqtu S ol-hasv va maqtu*} oz-zarb, p.v. 115. - ^ ^ w w — ^ -w- vazn e dhu ye kuhi or b. salim e *Jl* mn. mt. va sdlem va mx. va ahazz e nix. §258. ES 177, 186. (7.2. 2/3/ 1 / 2 / 2 / 2 / 2 ). v/- — b. basil e mn. mt. va maqtu *i (e mozdi } . ES 100. 3.4 6(2). - w - - w ^ — b. rajaz mn. e mt. os-sadr va-z-zarb ( el- mo zed) va maqtu*} va mx. ol-hasv , p.v. 96. v - w b. rajaz e mn. mt. os-sadr va maqtu*} va mx. ol-hasv va maqtu*} oz-zarb (et-mozdl), p.v. 96. rajaz e ms. mt. os-sadr va-z-zarb ( ei- mozdi) va maqtu 1 } ol-hasv, p.v. 98. ^ w — b. rajaz e ms. mt. os-sadr va-z-zarb el moraffal va maqtu*} ol-hasv, p.v. 99. - v ^ - - v ^ - b. rajaz e mn. mt. os-sadr va-z-zarb (el-mozdl) va maqtu*i va mt. ol-hasv, p.v. 100. — v — b. raja: e mn. mt. va maqtu*} (e mozdi), p.v. 100. - b. s. ms. mt. os-sadr va-z-zarb el-maksuf ( f-mouquf ) va maqtu*} ol hasv, p.v. 102. i w - w - 6. s. ms. mt. os-sadr va-z-zarb el-mo raffed va maqtu*} ol-hasv, p.v. 103. -o — b. badil e ms. mt. va maqtu*} va sdlem ( I mosbaq), p.v. 104. — rajaz e mo r abba*} e moraffal e mt. va maqtu*}, p.v. 1 14. - ^ ^ - -wv — b. rajaz e mn. mt. os-sadr va-z-zarb el-morqffal va maqtu*} e moraffal va mt. ol-hasv, p.v. 106, Or O O O' V o 1 o' o v> O 1 O' O' O^ O' O' ’ o O' 240 APPENDIX ONE 118 -w- 119 120 - w- 121 122 123 -w- 1 18 vw- w; b. raja z e mn. nit. va maqtu ¥ e moraffal. p.v. 106. b. rajaz e ms. mt. ossadr va maqtu V ol-haSv va-z-zarb ( el-mozdl ) , p.v. 98. w w — 'wv-- - v w — h. rajaz e nut. mo raff a! e mt. ossa dr va-z-zarb va maqtu V va mt. ol-hasv, p.v. 1 15. — w w — b. rajaz e mn. moraffal e mt. va maqtuS. p.v. 115. h. rajaz e ms. mt. ossadr va maqtu ol-hasv va-z- zarb el-moraffal, p.v. 99. Or (- w- — — ) b. basil e mn. mt. ossadr va maqtu V ol-hasv va-z-zarb (el-mozdl), p.v. 117. ^ ^ - -ww- b. rajaz mn. mt. ossadr va-z-zarb (ei- mozdl) vo maqtu V oi-hasv , p.v. 100. - v ^ — b. rajaz e mn. mt. ossadr va-z-zarb el-moraffal va maqtuH e moraffal va maqtu $ oi-hasv. p.v. 106. vu- b. rajaz e mn. mt. ossadr va maqtu *i oi-hasv va-z-zarb (el-mozdi). p.v. 100. ^ — -w'- b. rajaz e mn. moraffal e mt. os- sadr va-z-zarb va maqtu V oi-hasv . p.v. 115. b. rajaz e mn. mt. ossadr va maqtuS e moraffal va maqtu V oi-hasv va maqtu 5' oz-zarb el-moraffal, p.v, 106. w w — b. rajaz e mn. moraffal e mt. ossadr vo maqtu ? oi-hasv va-z zarb , p.v. 115. v - v b. x. mn. makfuf va mx. ( mozal ) or b. moqtazab e mt. va mx. (mozal). ES 117. (12.3.16). w - ^ \j — b. mosakel e morahba V e makfuf e mh. (JmqJ. ES 102. 4.2.07. \s — — — — b. mosakel e ms. makfuf e mh. ( Imq.). ES 102. (4.2.07/04). v- w v — v-v - b. m. ms. marfu ? e O' o o o 146 REFERENCE LIST OF METRES 243 - ^ ^ - w - v/ w - h. x. ms. sdlem os-sadr va mx. ol-hasv va-z-zarb el-mh. ((mq.), p.v. 20. - ^ — w - w - v w — b. x. ms. saiem os-sadr va mx. ol-hasv va-z-zarb ( el-mosbaq ) , p.v. 2 1 . -v — v - v - w — v - v - b. x. mn. saiem os-sadr va mx. ol-hasv va-z-zarb (el-mozal), p.v. 22. 135 -w — w - w - - w - b. x. ms. saiem va mx. va mh. ( /mq .), ES 115. (9.3.11) . 136 - v v v - - w — b. x. ms. saiem va mx. ( va mosbaq). ES 115-11 6. (9.3.12) . 137 - w — v v — b, x. mn. saiem va mx. ( mozal ) . M 2260. (9.3.16). - v- - w - w - — b. x. ms. sdlem va mx. va as. (mosbaq), p.v. 20. - w — w - v b. x. ms. sdlem va mx. va mosaWas ( e mosbaq), p.v. 21. -v — v/ - v- v — v - b. x. mn. saiem va mx. va mosaWas va mx. (mosbaq), p.v. 22. 138 -v- - u— - v- b. motaddrek e ms. sdlem (e mozal). ES 91. 1.3,09. 139 -v- b. m. ms. marfu ? va sdlem va mx. (mozal). ES 116. (9,8.11). 140 -v- -v— — v- - b. motaddrek e mn. ahazz ( e mozal). §190. Nlma Yusij: afsane. ES 91. 1.3.10. 141 -v/- - v - — ^ ^ - b. motaddrek e mn. sdlem (/mozal). ES 91-92. 1.3.12. 142 -v — ^ — b. mosdkel e ms. mh. (/mq.). ES 116. (10.4.11). 143 - v - - w — - w - — h. m. mn. marfu V va sdlem (/mosbaq) or b. motaddrek e mn. sdlem va moraffal ( /motavval). M 809, 1094, 1655, 2892. 1.3.07(2). (In divan e kabir poems in this vazn are less correctly ranged as b. r. mh. va sdlem (/mosbaq).) 144 - v — - w - b. madid e morabbab e sdlem. M 3197, ES 95. 2.4,07. (In divan e kabir M 3197 is printed as mosamman, but the rhyme arrangement clearly shows it to be morabbaS.) 145 - ^ — - v- — b. r. morabba 9 e sdlem (/mosbaq). ES 95. 2.4.08. 146 - v — - v - -v/ — -v — b. madid e mn. sdlem (e mozal). M. 237, 2086, 3109. 2.4.07(2). (M 3197 is a Greek molamma 9 exhibiting several irregular variations, e.g. - w — , w ^ ^ - w ~ and instead of - These have not been listed.) 244 APPtNDIX ONt 147 147 -u — - ^ — - h. r. ms. majhuf (e mosbaq). ES 95. 2.4.09. 1 48 - w — ^ - b. r. ms. mh. (jmq.). §§201*203. 2.4.11. 149 - vy — -« — - o — h. r. ms. salem (e mosbaq). ES 95. 2.4.12. 150 — - yy — - ^ ^ - b. r. mn. mh. ( \mq.) . §§ 199-200. 2.4. 1 5. 151 -w — - yy - - -yy — - w — h. r. mn. salem (e mosbaq). §198. 2.4.16. 152 - w — - w — - «y — — b. r. mn. as. (mosbaq). ES 95-96. 2.4.14 (sic !). 153 - vy — - yy — — b. jadid e ms. salem (e mosbaq). al -mo 5 jam 135. (11.8.4/8). -v — — -y v — ^ ^ - b. madid e mn. salem os-sadr va maqtu ‘i va nix. ol-hasv va m.x. oz-zarb ( el-mozdl ), p.v. 8. -vy — — w vy — — b. madid e mn. salem os-sadr va maqtu $ va ntx. ol-has v va maqtu 5 oz-zarb (el-mozdl ) , p.v, 8. 154 - yy — — w- b. x. morabba S e salem (e mozdl). ES 117. 11.8.08. - w yy - yy - vy v - />. //I. mn. marfuS va mosa Was va m.x. va mx. mh. (jmq . ) , p.v. 122. -yy- yy-yy- — 6. 777 . 777/7. marfu *1 va mosa Was va mx. va as. (mosbaq), p.v. 122. 155 - yy - — — vy — * — b. motaddrek e mn. salem va maq tu S (e mozdl). ES 95. 2.4.05(2). -yy — yy yy - h. r: 7775 . salem vq mosa 5 Sas va m.x. mh. (jmq.), p.v, 11. - yy — yy yy — b. t\ 7/75. salem va mosa^^as va mx. (mosbaq), p.v, 12. - yy vy yy yy yy - /?, 7\ 771/7. Salem V'77 7777/577 V Ws va mx. va mx. mh. (jmq.), p.v. 13. — * yy — yy yy — yyvy — b. r. 777/7 . salem os-sadr va mosa £ fas va mx. ol-hasv va mx. oz-zarb (el-mozal), p.v. 14. - y — y U — - b. r. mn. salem va mosa Was va mx. va majhuf (e mosbaq), p.v. 16. - yy yy yy 6. T. 77777. Salem VQ 777050 *1 ^05 VO 777.V. VO 05. (mosbaq), p.v. 13. - yy — yyv — 6. r. 77177. salem vo mosa 9 ^as va mx. va mosaWas (e mosbaq), p.v. 14. ■ - yy — yy — b. jadid e ms. salem va mosa^^as va mx. ( mozdl ) , p.v. 17. 159 REFERENCE LJST OF METRES 245 156 157 158 159 -- - w - - w - ^ ^ - b. jadid e mn. sdlem va moiaWas va mx. va m. x. mh. ( mq .), p.v. 18. - vj — ^ w - — h, jadid e mn . sdlem va mosa 5 fas va m.x, va as. { moshaq ) , p.v. 18. - v — — I — vu- h. madid e mn. sdlem va maqtu V va sdlem va m.x. (mozdl), p.v. 8. - v — -- -w — - - b. madid e mn. sdlem va maqtu 5 fe mozdl). M 1748, 3110. (8. 2.6(2)). - ^ — | - ^ — h.r. mn. sdlem va mosa V fas va sdlem va m.x. (moshaq), p.v. 14. - w — |~w — b. r. mn. sdlem va mosa 5 fas (e moshaq), p.v. 14. _ v — — h. r. ms. sdlem va mosaWas va as. (moshaq), p.v. 11. -v — h. r. ms. sdlem os-sadr va mosa 9 fas ol-hasv va-z-zarh ( el -moshaq ) , p.v. 12. -w — v, w - 6. r. mn. sdlem os-sadr va mosa Was ol-hasv va mx. oz-zarh el-mh. (el-mq.), p.v. 13. w — — h. r. mn. sdlem os-sadr va mosa 9 fas ol- hasv va m.x. oz-zarh ( el-moshaq ), p.v. 14. - w — h. r. mn. sdlem os-sadr va mosaWas ol-hasv va majhuf oz-zarh (el- moshaq), p.v. 16. -v — — h.r. mn. sdlem os-sadr va mosa 9 9as ol-hasv va as. oz-zarh ( el-mosbaq ), p.v. 13. - v — _ — h.r. mn. sdlem os-sadr va mosaWas ol-hasv va-z-zarh (el-moshaq), p.v, 14. - ^ ^ h. motaddrek e mn. maqtuS os-sadr va m.x. ol-hasv va-z-zarh ( el-mozdl ), p.v. 1. - ^ ^ ^ - — h. motaddrek e mn. maqtu 9 os-sadr va-z-zarh ( el-mozdl ) va mx. ol-hasv , p.v. 1. - vj v-v- v - b. h. ms. axrab e maqhuz e majbub (jahtam). ES 109. 5.1.09. This rhythm may well be considered p.v. 164, cf. § 1 56. v w - v/ - ^ b. h. mn. axrab os-sadr va maqhuz ol-hasv va sdlem oz-zarb (el-mosbaq) ES 110. (5.1.09/6). w w- v- w — h. h. ms. axrab e maqhuz e mh. ((mq.). §§252-253. 5.1.10. 246 APPENDIX ONE |#.0 R - - w ^ - vy - w — vy v - b. h. nut. a.xrab va maqbuz va makfuf va majbub (/ah lam), p.v. 166. — w w-v- v> — \y w — 6. h. mn. axwh va maqbuz va makfuf va mh. (mif.), p.v. 167. Cf. §260. 160 — .y vy - v- v — ^ w h. It. mn. as rah va maqbuz va makfuf va salem (jmosbaq). ES 110. (5.1.12/3). 161 — w w-w- w b. h. ms. axrah va maqbuz va salem (jmosbaq). §254. 5.1.11. R — vy >y-vy- vy b. h. nut. axrah va maqbuz va salem va abtar (fazall), p.v. 166. 162 — w vy-w- - h. h. ms. axrah e maqbuz e abtar (jazall), ES 109. 5.1.08, 163 — vy «y — b. h. morabba y vy — vy vy - 6. h. mn. axrah e makfuf e majbub (jahtam). §§255-260. 3.3.13. 167 — vy vy — v vy — vy vy — 6. /?. mn . axrab e makfuf o mh. (jmq.). §§220-222. 3.3.14. 168 — v vy — vy vy vy vy b. h. mn. axrab ossadr va makfuf ol-hasv va salem ( jmosbaq ) oz-zarb. ES 99. 3.3.15. 169 — vy vy — — vy vy b. fi. ms. ax rah os-sadr va makfuf ol-hasv va salem ( jmosbaq ) oz-zarb. ES 98. 3.3.11. R — vy vy — vy vy 6. h . mn. a.xrab va makfuf va salem va abtar (jazall), p.v. 166. — vy vy — vy vy — b. H. mn. axrab va makfuf va salem va ahazz (e mosbaq), p.v. 167. --vy vy — vy vy 6. h. axrab va makfuf va salem va moxannaq (e mosbaq ) , p.v. 168. 170 — vy vy — vy -vy- 6. qarib e ms. axrab e makfuf e mh. (jmq.). §158, GI 2, 52, 192. (4.7.02/08). (In GI 2 and Gl 52 this rhythm is an irregular variation of 55. — v> vy — vy - v- vy vy - 6, qarib e mn. axrab e makfuf e majbub (jahtam). ES 107. (4.7.02/11). 171 179 REFERENCE LIST OF METRES 247 172 — w w — w -w — b. qarib e ms. ax rah va makfuf va sdlem ( jmoshaq ) . §§241, 243, 245. ES 107. (4.7.02/09). — v v — v - v — - h. qarih e mn. axrah va makfuf va sdlem va ahtar ( fazall ), p.v. 171. 173 — v w b. h. morabha 5’ e axrah e sdlem {jmoshaq). ES 98. 3.3.07. 174 — h. motaddrek e mn. maqtu 9 va mx. (mozdl). M 24N 242, 1749-1750, 2092-2096. 2252-2253, 3136-3138, 3202. ES 98, 109. 3.3.05(2). This rhythm also occurs as p.v. 3. 175 — ^ w v - h. h. ms. axrah e tnajhuh ( f ah tarn ) . ES 118.(13.1 .09). 3 76 — ^ v — — v v/ — b. h. mn. axrah va nth. (fmq.). ES 99. 3.3.06(2). — -- — h. motaddrek e mn. maqtu 1 . r os-sadr va-z-zarh ( el-mozdl) va mx. va maqtitS ol-hasv , p.v, I and 174. Or ( — ^ w — ) h. h. ms. axrah va sdlem va ahazz (e moshaq), p.v. 165. R - - ^ v — v w - b. h. mn. axrah va sdlem va axrah va majhuh (j ah tarn), p.v. 166. — ^ w — ^ w - h. h. mn. axrah va sdlem va axrah va mh. (fmq.).), p.v. 167. 177 — w v — ^ ^ h. h. mn. axrah va sdlem (jmoshaq). §§223-225. 3.3.07(2). This rhythm may also be p.v, 168. — v w h. h. ms. axrah va sdlem va moxannaq (e moshaq ) , p.v. 169. R — v v -b.h. mn. axrah va sdlem va moxannaq va ahtar ( azall ), p.v. 166. Or (-- v ^ — - -) h. h. mn. axrah va mh. (jmq.) va ax ram va ahazz (e moshaq ) . p.v. 176. — w w — b.h. nut. axrah va sdlem va moxannaq va ahazz (e moshaq ), p.v. 167. — v w b.h.mn. axrah os-sadr va sdlem va moxannaq ol-hasv va moxannaq oz-zarh ( el-moshaq ) , p.v. 168. Or h. h. mn. axrah os-sadr va sdlem va axram ol-hasv va moxannaq oz-zarh ( el-mosbaq ), p.v. 177. 178 — w - h. basil e mn. sdlem va mx. (mozdl), GI 272, S 680 (T - ari , ddni ce goft), M 3114-3115, ES 117. (12.2.07(2)). — v- vv/- — w- — b. hasit e mn. sdlem va mx. va sdlem va maqtuS (e mozdl), p.v. 178. — w _ v _ h. mozdre 5" e morabha V e axrah e mh. (fmq.). ES 106. 4.7.06. 179 248 APPENDIX 0N£ 180 180 — w - w- w ^ — b. mozare V e ms. axrab e makfuf e mh. (fmq.). ES 106. 4.7.10. 181 — w - w - w « — v - b. mozare *i e mn. axrab e makfuf e mat mus ^ yAz* ( fmaslux ES 107. 4.7.12. 182 — ^ - v - ^ ^ ^ - v- b. mozare *i e mn. axrab e makfuf e mb. (jmq.). §§237-238. 4.7.14. 183 — w - ^ ^ ^ — b. mozare 9 e mn. axrab os-sadr va makfuf ol-hasv va sdlem (fmosbaq) oz-zarb. ES 108. 4.7.15. 184 — « -w-w w b. mozare V e ms. axrab va makfuf va sdlem ( fmosbaq). §239. ES 106-107. 4.7.11. 185 — v/- v-w- b, rqjaz e morabha 5 e sdlem va mx. (e mozdl). ES 106. 4.7.08. 186 187 . - w rajaz e mn. sdlem va mx. ( e mozdl ) . ES 108. (4.7.08(2)}. - — b. mozare e morabha 5 e axrab e sdlem fe mosbaq). ES 106. 4.7.07. , ^ w - w - v - b. r. mn. mosa Was e makfuf va sdlem va maskul va mb. ( mq . ) , permitted (or slightly irregular) variation of 4. — w ^ ^ - v — b. r. mn. mosa Was e makfuf va sdlem va maskul va sdlem (fmosbaq). permitted (or slightly irregular) variation of 5. — ^ ^ — b. r. mn. mosa^Sas e makfuf va sdlem va maskul va as. (e mosbaq). permitted (or slightly irregular) variation of 6. 188 — v ^ - w - b. mozare^e mn. axrab va mb. (fmq.). ES 107. 4.7.06(2). — ^ - v — — b. mozare *i e ms. axrab e sdlem e ahazz ( e mosbaq ) , p.v. 180. — vj -w — — ^ - b. mozare 9 e mn. axrab va sdlem va axrab va matmus ( / maslux ) , p.v. 181. — ^ -v- b. mozare 5 e mn. axrab va salem va axrab va mb. (mq.), p.v. 182. Or b. r. mn. mosa V Ws e makfuf va sdlem va mosa^Sas e makfuf va mh. (jmq.), permitted (or slightly irregular) variation of 4. 189 — ^ - w — — w - ^ — b. mozare 5 e mn. axrab va sdlem (fmosbaq). §240. 4.7.07(2). Or p.v. 183. Or b. r. mn. mosaWas e makfuf va sdlem (fmosbaq), permitted (or slightly irregular) variation of 5. ’07 REFERENCE LIST OF METRES 249 — v -v — — u — b. r. mn. mosaWas e makjuf va salem va mosaWas e makjuf va as. (mosbaq). permuted (or slightly irregular) variation of 6, — - vv — b. mozdre^ e ms. axrah va salem va moxannaq (e mosbaq ), p.v. 184. 190 — w- -wv- v - v- b. rajaz e ms. salem va mi. va nix ( mozal ) . ES 113. (12.3.1/5/2/4). 191 — w- - ^ - b. basil e morabba*! e salem (e mosbaq). ES 90. 1.2.07. 192 — -v- ^ — h, basil e ms. makbul (e mozal). ES 115. (9.2.10). 193 ^ b. m. ms. salem os-sadr va-z-zarb (el-mozdl) va makjuf ol-hasv or b. monsareh e salem os-sadr va-z-zarb ( el-mozdl) va ml. ol-hasv. ES 115. (9.2.12). 194 — -w — b. m. morabba *i e salem (e mosbaq). ES 91. 1.2.08. 195 — v ^ — v- b. motaqdreb e mn. af. mb . . (jmq.). ES 91. 1.2.10. 196 — w — ^ — w — b. motaqdreb e mn. al. (mosbaq). ES 9 1 . 1 .2. 1 1 . 197 — v- -vy- — w — -v - b. basil e mn. salem (e mozal). ES 91. (10.3.14). 198 — w- -v — — w- - v — b. m. mn. salem (e mosbaq). ES 91. 1.2.08(2). 199 — ^ b. rajaz e morabba V e salem (c mozal). ES 93. 2.3.08. 200 — v- — w ^ b. kdmel e ms. mozmar os-sadr va-l-hasv va ahazz oz-zarb ( el-mo sbaq ) or b. s. ms. maxbul oz-zarb el-maksuf (jel-mouquf) . ES 118. (12.2.04/07). 201 — \y- — w — b. rajaz e morabba *> e moraffal ( jmotawal). ES 93. 2.3.09. 202 — v- — \y — — \y — b. s. ms. ml. oz-zarb el-maksuf (jel-mouquf). ES 116. (10.6.11). 203 — w - — v- — w- b . rajaz e ms. salem (jmozdl). ES 93. 2.3.12, 204 — o- — — w - b. rajaz e ms. moraffal (jmotawal) . ES 94. 2.3.13. 205 — \y — — — v/- — b. rajaz e mn. ahazz (e mozal). ES 94. 2.3.14. 206 — — w - — o- — b. rajaz e mn. salem (e mozal). §§205-206. 2.3.16. 207 — w- — w- — w - — vy- — w - — w-6. rajaz e salem e esnd basara roknan. M 1785. 2.3.24. 250 APPENDIX ONE 20S 208 -- 209 210 - 211 212 213 W — w 214 215 - v - — v - — w- b. rajaz e mn. salem os-sadr va-l-hasv va maqtu oz-zarb (el-mozal). ES 94. 2.3.15. -v— — m b. rajaz e ms. salem os-sadr va-l-hasv va maqiuS oz-zarb (el-mozal). ES 93. 2.3.11. -v- — v ^ - — w — b. rajaz e mn. salem va mo raff al (fmotavval). E5 95. 2. 3.9(2). -v- b. rajaz e morabba 5' e salem va maqtu 5' (e mozal). ES 93. 2,3.07. — w — — v — b. motaqdreb e mn. al. va salem (mosbaq). §§184-188. 1.2.05(2). — v - — — b. basit e mn. salem va maqtuS va salem va mx. ( mozal ), p.v. 178. — w - — — w- — b. basit e mn. salem va maqtuS (e mozal), p.v. 178. — - - — w- b. rajaz e mn. salem va maqtu ? (e mozal). ES 94. 2.3.07(2). — w- v - v-w- b. rajaz e mn. maqtu va mx. va mt. va mx (mozal), p.v. 91. — w - w - - kj- b. s. ms. maqtu V va mx. va ml. maksuf (jmouquf), p.v. 93. — v- w- b. rajaz e mn. maqtu 1 -. va mx. (mozal), p.v. 91. — - w - — ^ ^ b, basit e mn. ahazz os-sadr va salem ol-hasv va mx. oz-zarb (el-mozal) or ( ^ ^ - w ^-) b. kabir e ms. mt. oz-zarb (el-mozal). ES 118, 186. (12.5.1/4/7). — „ ^ w h. h. mn. axram os-sadr ( va mosbaq oz-zarb). ES 93. 2.2.15, — — vw- w- b. motaddrek e mn. maqtu S os-sadr va maqtuS va mx. ol-hasv va mx. oz-zarb (el-mozal), p.v, 1. — -vv- w - ^ - -vw- b. rajaz e mn. maqtub os-sadr va mt. va mx. ol-hasv va mx. oz-zarb (el-mozal), p.v. 96. — - v v - w-v- b. rajaz e mn. maqtu ? os-sadr va-z-zarb (el-mozal) va mt. va mx. ol-hasv, p.v, 96. — - w v/ - -wv- b. rajaz e ms. maqtu*) os-sadr va mt. ol-hasv va-z-zarb (el-mozal), p.v. 98. — - ^ ^ — b. rajaz e ms. maqtu *i os-sadr va mt. ol-hasv va-z-zarb el moraffal ( fmotavval), p.v. 99. w “ W w “■ w W W MM - MM M M — M — M M M M M M W “ ““ M W 215 REFERENCE LIST OF METRES 251 W 'w> h. rajaz c mn. maqtu *i os-sadr va mi. ol-hasv va-z-zarb ( el-mozdl ) , p.v. 100. - -ww- - ^ w b. rajaz e mn. maqtu *} os-sadr va-z-zarb (el-mozdl) va ml. ol-hasv , p.v. 100. - -uu- -v-b.s. ms. maqtu*} os-sadr va mt. ol-hasv va-z-zarb el-maksuf ( jel-mouquf ) , p.v. 102. - - ^ - - w - s/ - h. s. ms. maqtu*} os-sadr va ml. ol-hasv va-z-zarb cl -mo raff a l ( jel-moiavval ) , p.v. 103. - - ^ ^ - - v> — b. badil e ms. maqtu *i os-sadr va mt. ol-hasv va salem (jmosbaq) oz-zarb , p.v. 104. - vv- — b. motaddrek e mn. maqtu*! os-sadr va-z-zarb ( el- mozdl) va maqtu*! va m.x. ol-hasv , p.v. I. - - w w - - - ^ ^ v — b. rajaz c mn. maqtu*} os-sadr va mt. moraffal va mt. ol-hasv va mt. oz-zarb el-moraffal, p.v. 106. - — -vw- b. rajaz e mn. maqtu *1 os-sadr va-z-zarb el -moraffal va mt. moraffal va mt. ol-hasv , p.v. 106. - _ w w b. rajaz e ms. maqtu*! os-sadr va-z-zarb ( el-mozd! ) va mt. ol-hasv, p.v. 98. - - wv- b. rajaz e ms. maqtu *} os-sadr va-z-zarb el-moraffal va mt. ol-hasv, p.v. 99. - - v ^ b. rajaz e mn. maqtu*} va mt. (mozdl), p.v. 100. - v v — -vv — b. rajaz e mn. maqtu *i va mt. moraffal, p.v. 106. - - v ^ b. rajaz e mn. maqtu*} os-sadr va-z-zarb (el-mozdl) va mt. va maqtu*} ol-hasv , p.v. 100. - - v v — b. rajaz e mn. maqtu *} os-sadr va-z-zarb el-moraffal va mt. moraffal va maqtu*} ol-hasv , p.v. 106. * ^ - b. basil e morabba *} e maqtu*! va salem ( /mozdl), p.v. 107. - - - monsareh e ms. maqtu*} os-sadr va mt. ol-hasv va-z-zarb (el-mozdl), p.v. 108. - - w - w w - - h. monsareh e mn. maqtu*} os-sadr va mt. ol-hasv va manhur (jmajdu*}) oz-zarb, p.v. 109. -w- -v- h. monsareh e mn. maqtu*} va mt. maksuf ( jmouquf) va m.x. va mt. maksuf ( jmouquf) (or h. basil e mn. maqtu*} va salem ( j mozdl ) va mx. va salem ( jmozal), p.v. 1 13. - ^ ^ — b. h. ms. a.xram e as tar e mh. ( jmq .), p.v. 159. 252 APPENDIX ONE 216 R- R R Or ( v - w — i b. monsareh e ms. maqtu 5 va mt. va ahazz ( e moshaq), p.v. 111. - - v ^ v-b. h. mn. axram va as tar va makfuf va majbub (ahtam). p.v. 166. - - ^ w h. h. ms. axram va as tar va salem ( j moshaq ) , p.v. 161. Or ( v - ^ ) h. monsareh e ms. maqtu V os-sadr va-z-zarh ( el-mozal ) va mt. ol-hasv , p.v. 108. - - ^ - w -A. monsareh e mn . maqtu C va mt. va maqtu 9 va manhur ( fmajdu < i), p.v. 109. Or ( -%>— ^ -) b. h. mn. axram va as tar va salem va abtar (jazall), p.v. 166. - - v - - w ^ - - w - b. monsareh e mn. maqtu*! va mt. maksuf ( jmouquf ) va mt. va mt. maksuf ( jmouquf ) (or b. basit e mn. maqtu V va salem (jmozdl) mt. va salem (j mozdl ), p.v. 113. - - v- - v - b. monsareh e mn. maqtu 5 va mt. maksuf ( jmouquf, j (or b. basit e mn. maqtu 1 ! va salem (jmozdl)), p.v, 1 13. - — b. rajaz e mo r abba ? e moraffal e maqtu V va mt., p.v. 114. Or ( — w « — ) b. h. ms. axram va axrab va mh. (fmq.), p.v. 165. ^ ^ - b. h. mn. axram va axrab va makfuf va majbub (j ah tarn), p.v. 166. - - - ^ m ^ w — b. h. mn. axram va axrab va makfuf va mh. (fmq.), p.v, 167. - — w w — v v b. h. mn. axram va axrab va makfuf va salem (/moshaq), p.v. 168. , - -- v ^ b. h. ms. axram va axrab va salem (jmoshaq), p.v. 169. - - - ^ ^ — b. rajaz e mn. moraffal e maqtu 5 os-sadr va mt. ol-hasv va-z-zarb. p.v. 115. - - ^ ^ — b. rajaz e mn. moraffal e maqtu 5 os-sadr va-z-zarb va mt. ol-hasv, p.v. 115. - — ^ w b, h. mn. axram va axrab va salem va abtar (jazall), p.v. 166. - — ^ v> — b. h. mn. axram va axrab va salem va ahazz (e moshaq), p.v. 167. - — w w b. h. mn. axram va axrab va salem va moxannaq (e moshaq), p.v, 168. R 216 REFERENCE LIS i OF METRES 253 — -w — b. rajaz e mn. mo raff al e maqtu*) va mt ., p.v. 1 !5. - — h. rajaz e mn. moraffa! e maqtu*) os-sadr va-z-zarb va mt. va maqtu *i ol-hasv, p.v. 1 15. — v -v/- b. qarih e ms. axram va axrah va mh. ( fmq .), p.v. 170 and irregular variation of 55. --w w- b. qarih e mn. a.xram va axrab va makfuf va majbub ( jahtam ), p.v. 171. — ^ ^ — b. qarih e ms. ax ram va axrab va salem ( jmosbaq), p.v 172, — w ^ — - h. qarih e mn . a.xram va axrab va salem va abtar ( jazall), p.v. 171. — — — motaddrek e mn. maqtu *) os sadr va-l-hasv va mx. oz-zarb ( el-mozdl), p.v. I and 174. — - ^ w - — b. basil e mn. maqtu *) os sadr va-z-zarb (el- mozdl ) va maqtu *) va mt. ol-hasv, p.v. 1 17. w - ^ - - w-b. rajaz e mn. maqtu *) os-sadr va maqtu V va mx. ol-hasv va mt. oz-zarb ( el-mozdl 7, p.v. 96. — v - v b. rajaz e mn. maqtu *) os-sadr va-z-zarh (el-mozdl) va maqtu 5' va mx. ol-hasv , p.v. 96. - w w - b. rajaz e ms. maqtu*) os-sadr va-l-hasv va mt. oz-zarb (el-mozdl), p.v. 98, -vw — h. rajaz e ms. maqtu *) os-sadr va-l-hasv va mt. moraffa! oz-zarb, p.v. 99. Or ( — — ^ ^ b. h. mn. ax ram va ahazz (e mosbaq ) va axrab va nth. ( jmq.), p.v. 176. _ u w w ^ - h. rajaz e mn. maqtu *) os-sadr va maqtu 5 va mt. ol-hasv va mt. oz-zarb (el-mozdl), p.v. 100, h. rajaz e mn. maqtu 5 os-sadr va-z-zarb ( el-mozdl ) va maqtu*) va mt. ol-hasv , p.v. 100. --- - b. s. ms, maqtu*) os-sadr va-l-hasv va mt. oz-zarb el-maksuf ( jel-mouquf) , p.v. 102. - w ^ - ft. $. ms. maqtu*) os-sadr va-l-hasv va mt. oz-zarb el-moraffal ( f el-mo tavval) , p.v. 103. -w — b. had'd e ms. maqtu*) os-sadr va-l-hasv va salem oz-zarb el-moshaq ), p.v. 104. 216 — — — — b. motaddrek e mn. maqtu*) (e mozdl). M 2097, 2418. (11.2.4(2)). Also p.v. I and 174. Or ( ) b. h. 254 APPENDIX ONE 217 ms. ax ram va moxannaq va ahazz ( e mosbaq ), p.v. 165. Or ( ) h. rqjaz e mo r abb a *1 c moraffal e maqtu *i, p.v. 1 14 (e.g. M 1362, line 14407). R — v v - b. h. mn. axram va moxannaq va axrab va majhub ( jahtam ), p.v. 166. — ^ w — fr fo mn axram va moxannaq va axrab va mh. (f mq .) , p.v. 167. - wv -- b. rqjaz e mn. maqtu 9 va maqtu 9 e moraffal va mt. va mt. moraffal , p.v, 106. — v v h. b. mn. axram va moxannaq va axrab va sahm f /mosbaq), p.v, 168 and 177. _ b. rqjaz e mn. maqtu 5 os-sadr va-z-zarb el-moraffa l va maqtu 5 e moraffal va mt. ol-hasv , p.v. 106. b . rajaz e ms. maqtu (e mozal), p.v. 98. Or h. It. ms. axram va moxannaq (e mosbaq ). p.v. 169. - w w — — b. rajaz e mn. moraffal e maqtu V os-sadr va maqtu S va mt. ol-hasv va mt. oz-zarb, p.v. 115. — b. rajaz e mn. moraffal e maqtu*i os-sadr va-z-zarb va maqtu 5 va mt. ol-hasv, p.v. 115, R b. h. mn. axram os-sadr va moxannaq ol-hasv va abtar ( j azai l ) oz-zarb, p.v. 166. Or b. rajaz e ms. maqtu ? e moraffal, p.v. 99. Or ( — ) b. h. mn. axram va ahazz (e mosbaq ), p.v. 176. Or b. basit e mn. maqtu ? (e mozal)), p,v. 1 17. -vw-6. rajaz e mn. maqtu S os-sadr va-l-hasv va mt. oz-zarb ( el-mozdl ), p.v. 100. — b. h. mn. axram os-sadr va moxannaq ol-hasv va ahazz oz-zarb (el-mosbaq), p.v. 167. Or b. ramal mn. maqtu V e ahazz (e mosbaq), p.v. 101. _ ^ v — b. rajaz e mn. e maqtu*/ os-sadr va maqtu*! e moraffal va maqtu ¥ ol-hasv va mt. oz-zarb el-moraffa! , p.v. 106. 217 b. h. mn. axram e moxannaq (e mosbaq), M 1494, 3183. 2.2. 3(4). Also p.v. 168 and 177. Or b. rajaz e mn. maqtu*r(e mozal), p.v. 100. — — — b. rqjaz e mn. moraffal e maqtu ? os-sadr va-l-hasv va mt. oz-zarb, p.v. 115. 217 REFERENCE LIST OF METRES 255 — — h. rajaz e nut. maqtuS va rtuiq tu ( i e morajfal , p.v. 106. — — b. raja r e nut. morajfal e maqtuS, p.v. 1 15. KEY TO THE BOHVR IN APPENDIX ONE axir, see mosakel . badil , 104. basil. 107. 113, 117, 178. 191-192. 197,214. hazaj, 33-34, 38, 40-41, 52-54, 57, 77-84, 134. 157-169. 173, 175-177, 215-217. jadid, 17-18, 153. kabir, 214. kdmei , 2, 3, 7, 200. madid, 8, 144, 146, 156. mojtass, 25-31, 44-45, 112, 121-124, 139, 143, 193-194, 198. man. sarcb, 36-37, 39, 48, 94, 108-111, 113, 193, moqtazah , 58-59, 125-127, 130, 133. mosakel , 119-120, 1 42. motadarek , I, 138, 140-141, 155, 174,216. motaqareb , 68-70, 72-75, 195-196, 212. mozareS, 49-50, 60-67, 85-87, 179-184, 187-189. qarih 32, 55-56, 170-172. qarib u; , see jadid. rajaz . 42-43, 46-47. 51, 88-92, 95-101, 106, 114-115, 128, 185-186, 190, 199, 201, 203, 211, 213, 216-217. ramal , 4-6, 9-16, 129, 131-132, 145, 147-152. salim, 116. sartf, 35, 93. 102-103, 105, 200, 202. lavil, 76. safer, 24. a aft/, 19-23, 118, 135-137, 154. APPENDIX TWO REFERENCE LIST OF SPFXLINGS The transcription used in this manual distinguishes all the Classical Persian phonemes and even indicates certain unphonemic niceties of pronunciation that are nevertheless prosodicafly significant, but graphemes with neither phonemic nor prosodical relevance are not distinguished. It is therefore not possible to tell the original spelling of words containing any of the letters h, q, s, t, z from the transcription. The following list gives the spellings of all words containing any of these live letters except such as occur only once since all words are given in the Arabic script as well as in transcription at their first occurrence. abtar Jj>) abydt Owl afsdne aII-jI ahazz X*-' Ahli ye Slrazl lf*' Ahmad -t-w*-' $ At tar jUaP A xa van e Safes azall Jjl Azraqi Jjj' Baba Taker j *' IX Bahai (JV*. ah tarn dhu ye kuhi \Sy* Saruz Asacfi as lam 1 ' aslam 1 ( J-* ? ' Asliani bahr j£ basil -X. behest OX,- v » m beit - » bohur bits tan Daqiqi l In Appendix One P is abbreviated at. and ^W>i is abbreviated as. In all other occurrences the Arab spelling has been added so as to keep the two homonyms distinct. 258 APPLNDIX TWO Dehla \i deraz j'^P ebtedd Eqbal eqbal name eskandarname 4^^ esnd ‘iasara roknan E^tesdml ezdfe ■ Ferdousi Foruqi Jr)} Foruzdnfar Ji^j)} golesidn jli—iS" Hafez hamze hasa vein ^ hast hasv hazaj Helali J*>U Homal JU* liahdhddi t£jW . * ** J loi MostoujT motadarek motaharrek motaqdreb nwtavval djz* mouquf moxannaq Dchlavi :olei\a REFERENCE LIST OF SPELLINGS 259 mozal jl-i* mozare ? mozmar Ndser Naslr od-DIn e TusJ nazm [*-» Nezaml nJm-fathe OuhadJ ouzan qarib 1 <—i j* O' jy qasd'ed JJwa» qaslde qazaliat OU Ijp QanJ ^ Qazvtnl Qeis 0 r-r , Qozdart 1 VOV f W/ Ouaf- jlj vozn Ojj xamse Xdqdnl A'o.yrou f *—*>- yusof zafarndme *A>j& Zdkam ci^Li zarhein uy.j*? zoieixd 2 In order to avoid confusion jadid has been employed instead of qarib v;j*-Cf. §244. SUPPLEMENTARY NOTES Ad §27 In this paragraph it is stated that the vazn e motaqareb “is probably the rhythm most frequently met with in Persian poetry". This is of course a perfectly subjective judgement. How often a particular reader meets with this or that rhythm, depends upon what type of poetry and which authors he reads. It is therefore not possible to say which rhythm is the one most frequently met with. One could have recourse to statistics. El well -Sutton (pp. 145-167) has an interesting Statistical Survey of Use of Metres. According to him the twelve commonest metres are: 1. 4.1.15 i bahr e mojtass §§ 228-229) 2. 4.7.14 {bahr e mozare 9 §§ 237-238) 3. 2.4,15 bahr e ramal e mosamman §§ 199-200) 4. 3.1.15 (i bahr e ramal e maxbun §§211-212) 5. 4.5.11 (bahr e .vu/f/ §§230-232) 6. 2.1.16 (bahr e hazaj e mosamman §§191-1 93) 7. 3.3.14 ( bahr e hazaj e axrab e makfuf e mahzuf §§ 220-222) 8. 2.1.11 (bahr e hazaj e mosaddas §§194-1 96) 9. 2.4.11 (bahr e ramal e mosaddas §§201-203) 10. 5.1.10 (bahr e hazaj e axrab e maqbuz e mahzuf §§ 252-253) 11. 4. 7. 7(2)' (bahr e mozare 9 e axrab va salem §240) 12. 1.1.11 (bahr e motaqareb §§ 178-183) It is seen that vazn e motaqareb occupies only the twelfth position. According to Elwell-Sutton (p. 162) it accounts for only 1.9% of the total. However, in Elwell-Sutton 's statistics a poem is a poem and length does not count. sahname does not carry more weight than any short qazal. Interesting though they be, these statistics cannot tell us which rhythm is most frequently met with. In a statistical analysis of the entire Persian poetic literature, counting the actual number of lines occurring of each rhythm, the ma.snav/-metres would certainly move up and vazn e motaqareb would probably top the list, but this too would not decide which rhythm is the one most commonly met I Elwell-Sutton has 4.7.15. obviously a misprint, Ad $27 SUPPLEMENTARY NOTES 261 with, for in this analysis the 75000 ah y a i of Mostoufi's zafanmme would outweigh the combined poetic production of SaSdi, Hafez and Moulavi, and zafarname is preserved only in a single manuscript, so it cannot be said to be commonly met with. In this connection it would not be without interest to examine the relative frequency of the various metres occurring in works where prose is mixed with poetry, for in such works the poets seem to move more freely between the various metres than in other types of poetry. I have therefore made a counting of Sa c di's golestan. Here vain e motaqareb stands fourth on the list, but this, too, cannot of course decide the question. The results of my counting are as follows; Rhythm number Elwell-Sutton's Number Total number in Appendix One code-number of poems of lines 20 4.5.11 159 337 80 2.1.11 77 163 28 4.1.15 77 132 73 1.1. 11 70 111 166 3.3.13 46 55 182 4.7.14 30 56 159 5.1.10 23 52 13 3.1.15 23 39 102 3.4.11 21 43 148 2.4.11 17 37 167 3.3.14 14 21 150 2.4.15 8 11 109 4.4.13 6 7 1 1 3.1.11 4 8 113 4.4.07(2) 4 7 74 1.1.12 2 3 84 2.1.16 2 2 151 2.4.16 2 2 178 12.2.07(2) I 2 Irregular rhythm: — — — * u — “ — — \j — 1 2 5 5.3.16 I I 14 3.1.16 1 1 262 SUPPLEMENTARY NOTES Ad §27 76 1.1.07(2) 1 l 2 91 5.2.J6 1 1 161 5.1.11 1 1 165 3.3.10 1 1 168 3.3.15 1 1 206 2.3.16 1 1 Ad §38e fn. 3 The explanation alluded to in the footnote explains pir e mard as “vieillard male", i.c. considers pir a noun and mard an adjective. However, I think it can be shown that this explanation is incorrect; If we add the indefinite \ to pir mard we get pir mard i (cf. bozorg mard /), but if we add the indefinite i to pir e mard we may get pir i mard as well as pir e mard i (cf. mard i bozorg and mard e bozorg /'). Now, in Classical Persian is quite common, but I do not remember that I have ever come across Therefore the interpretation pir e mard "vieillard male” can hardly be correct. Ad §72 in. 19 Hubert Darke quotes 24 examples of -ee, and four examples of -eo. I am not aware of any native tradition for the pronunciation of -eo, but on the analogy of -ee to be pronounced -ei ( <-aj ) it would be reasonable to pronounce -ej> as -ou { < -au). Then he gives eight examples of -ie. Both in Classical and Modern Persian recitation this should be pro- nounced -i (from Classical Persian -/ / with final -/ shortened before ezdfe , cf. §89). Finally Hubert Darke gives two examples in which the ezdfe cannot possibly have been contracted with the preceding syllable, namely sux'tan e kuh to be scanned - ^ — (his no. 6) and axer e sab az to be scanned -- w- (his no. 12). In the last example we can certainly read axer sab “late night" instead of axer e sab. sux'tan e kuh is more difficult to explain. Probably the poet has here allowed the infinitive to take a direct object without ezdfe as a finite verb would do: sux'tan kuh. I therefore cannot agree with Hubert Darke in “supposing that in this single instance Classical Persian prosody allowed two light syllables to take the place of one heavy one”. 2 This is a molamma ¥ with permitted variations following the rules of Arabic prosody. Ad $208 SUPPLEMENTARY NOTF.S 263 Ad §86 fn. 5 In Elwell-Sutton’s material Balyuzl has three occurrences of - u:n , namely cun 50 (p 200), cu:n 62 (p. 203), and xu:n 35 {p. 204), Hence the average length of his -u:n must be 49 (hundredths of a second), not 63. Ad §92 In transcribing mian instead of meyan or mi van I follow Lazard (/wdn,Grammairedupersancontemporain,§2l)and Lambton (mian, Persian Vocabulary, Cambridge 1961, p. 162). Ad. §124 The exact relationship between pesar and pos is as follows: The genuine Persian word is pos, which goes back to Pahlavi pus from Old Persian puqa- ( A vesta n puOra). Pahlavi had two words for “father", pid from the Old Persian nominative pita and pidar from the Old Persian accusative *pitaram. Pahlavi pusar arose beside pus on the analogy of the pair pid; pidar (and similar pairs: mad/ madar “mother", dux ij dux tar “daughter”). Modern Persian pesar for posar has been further assimilated to the word for father: pedar. (Modern Persian pur “son" goes back to Parthian puhr.) Ibidem That the Russian editor reads auramazd instead of urmozd is clear * from the index to Vol. VIII, p. 8. Ad §150 fn. 10 The verse quoted by al-motyam is: Jr 4 0l>- hezanamat bebaramat jdn c man 1 hegazamat hemazamat bar do tab ^ ^ ^ ^ -) “Let me hit you and carry you ofT, o my soul, let me bite and suck both your lips". (A wonderful flight of imagination !) Ad §178 “The most well known...". Cf. Ad §27 above. Ad § 187 One occasionally meets with poems in which some abydt are doubled and others are not. Thus in M774 line 8072 is not doubled whereas line 8080 in the same poem certainly is doubled. (Most lines in this poem admit of both interpretations.) Ad §208 fn. I The persicus or as it is normally called iontcus ^ — ) appears to be a foreign element in Greek prosody: “Die unmittelbare Folge zweier Longa innerhalb eines Metrums ist in der griechischen Metrik ohne Parallcle, so daO die Vermutung, der Ursprung des lonikus sei auBerhalb des Griechischen zu suchen, naheliegt, ... Dieses VersmaB diente aber auch der Charakterisierung des Oriental isch-Fremdartigen, weshalb es besonders in den 'Hiketiden’ und ‘Persern' des Aischylos und den Bakchen' des 264 SUPPLEMENTARY NOT'-S Ad § 208 Euripides vorherrscht" 3 . It seems thus, that apart from the statement of the Alexandrian scholiast there is also internal Greek evidence which points to a Persian origin of this metre. Ad §215 Professor Annemarie Schimmet suggests in a letter that the reason why this vazn is common in Moulavi and rare elsewhere “is probably that it is a typical dancing metre and therefore very fitting for poems recited during the whirling dance". Ad §231 bahr e xajif is especially common in the type of poetry known as qerfe. Typically more than a quarter of the verses in golestan are in this vazn. Ad §251 fn. 4 The beit referred to is p's. Wi ^ -A 1 Here the rhythm shows us that we must read .vdniam “you call me" and berdniam “you chase me away" (not X'anim and beranim). Ad §260 It is interesting to note that the fard is much more common in golestan than the robaSi proper. There are 37 instances of the former and only 9 of the latter. Ad §320 fn. 51 The etymologically correct pronunciation of the obsolete Off, kau or kau (cf. Old Hindi kahu and kahu), has been "lost” and native scholars nowadays pronounce it ku. This too is the pronunciation given in the glossary of Matthews and Shackle: An Anthology of Classical Urdu Love Lyrics* but the correct pronunciation is still found in John T. Platts: A Dictionary of Urdu , Classical Hindi , and English , from 1884. 3 Dietmar Korzemewski: Griechische Metrik, pp. 1 16-1 18 (Darmstadt 1968). INDEX Prepared by G re the Thiesen, cand. scient. The references are to paragraphs. The list of afaSU in §189 is not included in the index, h. stands for hahr e. In the alphabetical order x and .r" are reckoned as one letter. aha II, 121 ahag II, 121 abar 1 2 1 abe 121 abgandan 1 22 abi 121 m Abraham 339 1 - 1 abru 1 22 ahiar 255 Abu ... see also Bu ... Abu Bakr 21 5 9 Abu Hafs e Socjdi 258 abvat 24 Adib Saber 213 afd Si/ 189 dfarin 38 afarinname 1 8 1 aferidun 1 2 1 afgarukm 1 22 Afghan pronunciation of Persian 5 ajkandan ! 22 afqan 1 07 afraz 1 22 afruxt 1 22 afsdnd 1 22 afsdne 122, 190, Appendix One: 140 aftord 1 22 afsun 1 22 a/sils 122 Safu, fafv 124 afzud, a fzun 1 22 agar 60, 60' 2 , 120, 123 dhanin(e) 121 aherman 1 1 9 AhlTycSirazT 170.195, 205 Ahmad SAIT Ahmad 179 7 Ahmad e Tabriz! 179 ahreman, ahriman 1 1 9 ah ram 255 dhu ye kuhi 258 w m Sain 52 aind 318 dine ye sekandari 153 19 , 180 SAiS 284 Akbar Ilahabadi 306. 31 1 alef e tnaqsure 127, 146 alldh 127 al-mo Sjam 107, 107 s , 150 1 °, 162, 243 3 \ 255 1 , 258 13 , Supplementary Notes Ad §150 amir 1 20 Amir Xosrov e Dehlavi 153, 195, 218, 232, 250, 253, 264* andr 1 20 dnc ( e } 82 anceps 41, 43. 76, 166 Andrews, Walter G. 334 1 angizdnad 1 1 8 an is of- Sarefin 202 dnk f e) 82 anuse 1 2 1 m ■■ Anvari 212, 254 ar 60, 60' 123 dr 123 266 INDKX Arabic Arabic 34, 52. 54\ 59. 94, 104, 110, 111,125,126, 127. *45, 21 0 4 , 217, 241, 261. 262, 331 drad 1 23 drdmid 1 18 Arat, Re$id Rahmeti 324 2 , 326 s , 327® SArefi 253 arkdn 24 arqanun 1 68 1 1 Saruz 9, 25, 34, 177, 177 5 , 185 Asadi 68 astam 185, 228 4 aslam ‘i-l 220. 228 4 dsmdn 38, 296 dsoftand 119 asrarndme 1 95 ast 47, 77, 78, 98, 99, 100 astar 254 Astiani 127 2 , 175 asuftand 1 1 9 9 atabat ul hakdyik 333' 1 SAttar 191.195,198,202,266 aur 285 Aurangzeb 257® a\a 123 avartad ) 123 avardan 147 3 avaridan 121 dvdz 123 Avestan 66, 118. 124, Supplementary Notes Ad § 1 24 dvordan 121, 1 47 3 Axavan e Sales 167. 237 b. axir 244, cf. mosdke / axrab 220, 237 axram 239, 254 c Ayyuqi 1 82 azdd 1 20 azdd, seSr e 9 azail 255 azird 1 2 1 AzraqT 1 1 1 bd 11, 121 Baba Tahcr e SOryan 194, 196 hated, hae(c)e(gani. baegdn 66. 67, 70, 111, 295, 295 2N bdd( a) 121 baddye 9 175 b. badil Appendix One: 104 hadtar 1 1 1 Bahai, Seix 172, 202. 209 Bahar I 96 4 bah in, bahan , bahn 282 hahr 27, 170, 176, 267, Appendix One: Key to the bohur Bah ram e Gur 232 Bahrain e Pazdu 179 Bailey, H.W. 66 10 Bailey, T. Grahame 273 6 , 295 28 , 310 38 Baki 335,340,341,343 bar 47, 48, 121 baru 1 22 b. basil 261, 267, Appendix One: 107, 113, 117, 178, 191, 192, 197, 214 baft) tar 111 be- bo- 80, 93, 102, 103, 104 be 16, 70, 80, 81, 104, 124 bed 124 bed 24, 25, 26. 27, 185, 187 Bertels 117, 122 18 , 124 be run 1 1 8 bexrad 1 06 * W'fajram 283, 284 bi 121 Bidel 182, 195, 232 bihode , hihude 1 1 8 i f f birun 1 1 8 bo- 103 bodfam) 116, 118 bohur 170' 5 , 176 boland 6, 35, 41, 42, 44 borou 96 horun 1 18 Boulton, Marjorie 263' Browne. E G. 4, 26 2 , 141, 144, 219, 231, 257 7 , 276 7 budfam) 116. 118 fasunv INDLX 267 Bu SaSide bne Abe l-Xeir 156 Bu Sakur 59, 181 husidn 27, 49 6 . 98. 178. 181 caesura 186, 187, 223, 235, 240, 306 Qagatay, Saadet 327® cahdr 1 23 cap(p) 112 car 123 ce 65, 70 cefhejl 123 cenad 1 1 8 cerd 70 da, de, deft 293 cinad 1 18 cist 99 city 303 co 75 comer d 86 con 87 Cooper, J.R.S. 228 3 cun 75, 87 da ere 159-172.243.244 Daqiqt 54, 71, Appendix to Part One 1 q dar( r)andegdn 112 Dard 296, 297. 306, 316. 319 Darke, Hubert 72‘‘\ Supplementary Notes Ad §72 das( t) b(e) ah 81 degar 116, 118 Dehlavi see Amir Xosrou deram 1 07 derdz 35-43 Devellioglu, Peril 335* d b ar(a)m 283. 284. 290 digar 65 s , 116, 118 divan 1 75, 264-267 divan e kabir 39, 39 4 , 175, 267 divan i hikmet 333 1 1 * divan i fugat it lurk 333 1 1 do 74 dobeiti 141 dodigar 65 5 donyd, donyi '46 doxtar e raz 222 Ebn e Dorosluye ye Fasal 107 Ebrahim 339 13 ebtedfi 185 14 , 242 h 316, 317 eldhiname 1 95 Elwell-Sutlon, L,P. 12. 12 s , 28, 37, 38 J , 51, 51". 52'\ 76, 76“ 85, 85 s . 132, 135®, 139. 156 27 , 163, 165, 172' 8 , 186. 187, 245, 255 3 . 257®, Supplementary Notes Ad §27, Ad §86 emdie 145, 146, 238 26 , 337* Eqhal. Mohammad 196, 202, 235, 312 eqbaindme 1 55 Serf ion 232 eskandar 1 22 eskandamame 155 22 . 180, 180' 0 eskaram 1 22 csnd Sasara Appendix One: 51. 207 espahbpd 1 22 espar 1 22 espordan 1 22 esidb 122 estadan 118 E^tesami. Parvin e 127 Ethe, Hermann 1 79 7 ezdfe 22, 38, 47, 54-61, 63, 71, 72, 76, 84. 89, 1 14, 292, 336 fa Sal 32 et passim faSata 32, 189 Faiz 296, 296 30 japan I 07 fardmos , faramus 1 1 8 fardz 1 22 fard 260, Supplementary Notes Ad §260 far(r)ox 1 1 1 Farroxi ye SistanI 147, 148, 149, 229 faruxt 122 Jarvaf r } din 1 23, 1 24 Farzad, MaSSud e 256 fasdne I 22 268 INDEX fait e ihpur fail efhpur 283 fa^iti 32 et passim fashion 32 et passim Faxr od-DIn e Gorgani 195 Jegandan l 22 FeizT 229 s feraqname 182 Ferdoust 27, 31-127 passim, 131, 1 172 Ex., 178, 179, 182 fertdun 121 fesatui 1 22 fesord 1 22 fetddan i 22 Firth, JR 273 6 , 310 J * fonun e haldqat o sendSdt e adabi 129 1 , 144 Foruqi I27 2 , 175 foruxt 122 Foruzanfar 39 4 , 175, 267 fosun 1 22 Josus 1 22 fozitd, fazun 122 French 289. 289- 6 . 290, 290 27 frusad 107, 147 Galib 284, 298, 300, 304, 31 1, 313 Gal ib Dede 343 ganj c xo.xan 175, 259 gar 60' 2 , 120 gdv 118 Ghosh, J.C. 268 gid(h) 123 Gibb, E.J.W. 141, I41'\ 143 14 , 144 16 , 325 s , 326*, 339 13 gohar l 1 8 gales tan , go/setan 3, 108, 109, 143, Supplementary Notes Ad §27, Ad §231, Ad §260 go! o nouruz 1 95 goisan e raz 195 Gorganl, Faxr od-Din e 195 gorizandan 1 1 8 gorosne , go r sene 108, 109 gou 1 1 8 gouhar 1 i 8 gova(h) 123 Greaves, Edwin 288 Greek 128 s , 200 8 . 208 f , 263, 267\ Supplementary Notes Ad §208 guy o cougan 253 hadiqat ol-haqiqat 232 Haenisch, Erich 128 4 Hafez 43, 50, 51, 57, 76, 118", 154. 166, 175, 177, 182, 187, 192. 199, 212, 223, 229, 246 haf t peikar 232 Halil tir Rahman 339 1J hdhtdme 253 hamasesarai dar Iran 1 79 7 * v hamza. hamzg 2 1 . 30, 44-53, 9 1 , 287-292 harc( e) 82 hareka ( 126 Harley. A H. 273 6 , 3I0 J8 hasavein 185 HasimSAIi 318 Hasrat 280 hast behest 153, 232 hasv 185 Hatef 231 Hatefi 1 79 hour 285 b. hazaj 176, 191-197, 220-225, 252-260 hazaliat 175, 231 12 hazmdn 1 24 hedie, hedye 1 24 heildjname 1 95 Helali 195, 232 Henning, W.B. I30 2 Hephaestion 208 1 he sab e Saruz 34 hie 1 23 Hindi 38, 60' J , 67, 86, 268, 268 281, 288, 289, 289 2S , 290, 294, 295, 303, 310 hindl 268'. 323, 323 52 Homai, Jalat I29 1 , 144 homdy o hotndyun 1 82 Lval], Sir Charles INDEX 269 Homer 263 3 hormoz fd) 1 24 Horn, Paul 4 has 1 18 Hosein 238 host var 1 24 host! e nuiqta t i/matla V 144' 6 hosydr 1 18 hova , hu 126, 296 32 hus 1 18 ■ husidr, husvar 118, 124 ic 1 23 i- lif-ar 318, 319 ik 317 llahahadt, Akbar 306,311 imdle 337-341 mc( e) 82 Indo-Persian 5, 37, 38 ink(e) 82 Iqbal, Sir Muhammad 196. 202, 235, 312 fraj Mirza 252 istadan 1 18 A tzdfd 292 izafet 336 Jackson, A, V. Williams 66 n , 1 1 8 1 0 b. jadid 244, 245, Appendix One: 17, 18, 153 jam e jam 232 Jami 172 Ex, 195, 202. 214, 218 , 232, 253 jamsfd o x" qrsfd 1 95 Jan (d) var 87 Javan rrnrd 38, 86 /avidndme 202 Junker/Alavi 4 b. kabir Appendix One: 214 kali/e ho demne 202 ■■ i m b. kamei 261, 262, 267, Appendix One: 2, 7, 200 Karakhanidic 324-334 kdsk 39, 233 kaste re ezdfe 54 : , see also ezdfe kasre ye momd! 58 kau . kau h 320, Supplementary Notes Ad § 320 kavani 96 ke 65. 67. 70, 79 Kent R.G. 66 li , 86* 2W kermdn 276" key ani 96 ki 274. 295, 303 kiani 96 kid'ar 3 1 9 kij(i)e 313 Kiseleva, L.N. 5, 276 7 kiya 280 Klima, Otakar 208' koh 118 koi 304 kolliat e sa^di 1 75 ko/lidt e Sams 39, I00 17 , 128 s , 141 1J , 189, 267 Koprulii, M. Fuad 325\ 326 7 Korzeniewski, Dietmar Supplementary Notes Ad 208 3 ku Supplementary Notes Ad §320 kuh 1 1 8 kutadgu bi/ig 139, 324-333 kuldh 6-8 k'd 280, 314 Lambton, A.K.S. 4, 21, 54', 257*, Supplementary Notes Ad §92 Lazard, Gilbert 7, !6\ 38, 40. 77, 78. 80 2 *, 93, 99, 101, 105, 1 10, 121, 123. 256 s , 258, Supplementary Notes Ad §92 leild, leili 1 27 leild i vo (leili yo) majnun 140, 194, 253 Lesano-llah 257® Lewis, G.L. 335 s liken 145 LotfSali Azar 195 Lyall, Sir Charles 268 270 INDEX Mackenzie, D.N, MacKenzie, D. N. II 3 b, madid 261, 267, Appendix One: 8, 144, 146, 156 Mahmud e Qaznavi 172 Ex, 229 Mahmud c Saba 236 Mahmud e Sabestari 195 Mahmud i Ka$gari 333' 1 mahrmtd o ayydz 1 95 mahzuf 27,* 42, 176, 194, 199, 217, 228, 237 majbuh 255 majdu 5' Appendix One: 39 majhuf Appendix One: 16 majhuf 2, 5, 18, 131, 145 majnun o let la (leiiij 140, 194, 253 makbul Appendix One: 31 makfuf 220, 237, 239, 24! maksitf 217 Maktabi 253 malfuz 61, 85, 91, 151 manhur 233 Mani, mdni 140, 140 v , 146 Mansur e Manteqi 241 Mansuroglu, Mecdut 333' 1 numteq oi-teir 202 Manucehri 222 maqbuz 252 maqsur 42, 176 maqsure 1 27, 1 46 maqta V 1 44 1(1 maqtu ¥ 233 mardsi 175 marbu ( f Appendix One: 15 marsie 200 trnSruf 2, 5, 13! Marvazi. MasSudi ye 195 masdrft 24 maskul 248 maslux Appendix One: 181 masnavi 138, 139, 142, 153, 172, 178- 183, 194-196, 196\ 201-202, 209. 213- 214, 214", 216-219, 230-232, 239, 253, 263, 264, 326 7 mamaviat 138, 175 masnavi yv maSnavi 51, 58, 202 MasSudi ye Marvazi 195 mafia ¥ 144 lto , 154 matmus Appendix One: 181 Matthews, D.J. 277, 323 s 1 , Supple- mentary Notes Ad §320 mam 215,217,226.233.246,250 maxbul Appendix One: 200 maxbun 209. 226. 227, 244, 250 maxzan obasrar 218, 219 mazhar 1 95 Mazhar 305 mazhar ol~ Sajaeh 202 McGregor, R. S. 286 11 *, 288 mehtndn 1 ! 9 merd 316, 317 merad 1 1 8 mesbdh ol-arvah 253 Mesihi 339 mesrd o sakar 1 72 sobhat ol-abrar 214 soboldy 1 28 somdri 1 1 8 Somnath 229 so tor 122 Soz 314 Steingass, F. 4, 1 72 19 sub(a)h 284 Subutai 128 Siileman Dede 344 Ex surnndt 229 su- ve 97 syllabic poems 267, 326, 326 7 , Appen- dix Three id i l mM’ e sbfyri 9 Tafazzoli 266 tahamtan , tahmotan 1 08 7 tajvid 1 25 1 tamumame 1 79 ■m tanvin 126 tasdid 9 1 ta-tan-tan 33 taqazzol 141 tar(a)h 311 tarane 259 tarjiSband 144, 231, 253 tarkibband 1 44. 238 tarsd, tarsi 1 46 b. tavif 261,267, Appendix One: 76 lay y ebb t 175, 234 tazkere ye haft asmdn 179' tazmin 1 78 teimurname 1 79 ■ Tekin, Talat 333 11 telesm e heirat 1 95 terd 316, 317 Thomsen Hansen, KAre 128* lira 316-317 to 74, 99 tohfat ol-ahrdr 218 toilful olSerdqein 253 tur e ma Srefat 1 95 Turkish 37, 38, 40, 86, 111 10 , 120, 1 39, I44' 6 . 324-344 uftddan 11, 119 timid 1 19 Urdu 86, 268-323 urdti 268 1 urmozd 1 24 ■» us tad 1 19 ■* 274 INDEX Ovcysi Oveysi 344 vcl 309, 310 vafa 285' 7 b. vafer 261, 267, Appendix One: 24 vagar(na) 60, 81 w?', vahd 309, 310 vak‘ 300 Vali 278, 284, 285. 28 5 17 , 305, 308 vali, valik(en) 145 var 60 va ra 118 varna 60, 295, 303 varqe wo golsah 182 vase/at ol-Seqd 231, 253 vdv e *iatf 54-60, 63, 292, 336 vav e tnajhui 2-5, 18, 131, 145 vav e rnaSruf 2, 5, 131 va: 60 vazn 27, 1 76 vei 1 1 8 vis o rdmin 195 vo 300 Wolff, Fritz 4 ,v" dbdnid 1 1 8 b. xaflf 230-232, 237, 267, Appendix One: 19-23. 118, 135-137, ^.Supple- mentary Notes Ad §231 X w aju ye Kermani, 182, 195 Xalil or-Rahman 339 13 xamos l 1 8 xamse 153, 219 xarnuS 1 1 8 Xanlari, P.N, 163, 172' 8 . 214 Xaqani 212, 253 xavatim 1 75 xeradndme 1 55 Xosravi 1 58 Xosrou sec Naser Xosrou xosroundme 1 95 .vosm v o sirin 195 yd 309, 3 1 0 yd ' e majhut 2-5, 18. 131, 145 yd v maSruf 2-5.131 yahd 309, 310 vab 1 300 ■w Yaqma ye Jandaqi 210 ye 300 Yusij, Nima 190, Appendix One: 140 yusof o zoleixa 182. 195 Yusuf Ulua Hacib 324, 331, 333“ zdd 120 zqfamdme ye mostoufi 128, 1 79 , Supp le- mentary Notes Ad §27 zardiosindme 179 * zarb(ein) 177, 185 zardbest t :ard(h)osi 124 ze 70, 79 zenhar, zinhdr 87 zird 8 1 , 86 Zohurl 182 Zolall 195 zo l-bahrein 1 70 zoleixa 182, 182' 3