RHETORICAL FIGURES (ṣanāyeʿ, ṣenāʿāt, “artifices;” badāyeʿ “novelties;” maḥāsen, moḥassanāt, “beauties”), devices of embellishment, tropes, and figures considered as an intrinsic part of literary expression in medieval Persia (on the taste for the use of rhetorical devices and ornament as one of the general features of Persian literature see Yarshater, 1988, p. 18). From the late 10th century, the embellishment of speech (soḵan, kalām) gradually assumed a primary importance for both poets and scribes who were eager to delight and impress their audience and create verbal marvels. A century later the art of applying rhetorical figures became almost synonymous with literary craftsmanship. ʿOnṣor-al-Maʿāli Kaykāvus b. Eskandar in his Qābus-nāma (1082) recommended the use of figures (ṣenāʿāt) and figurative language as a proven path to the heights of eloquence, because in his view unembellished “plain or straight” verse (šeʿr-e rāst) lacked elegance (Qābus-nāma, p. 189). The Persian tradition of adorning and beautifying a poem in the manner of a “bride,”ʿarus, a topos of classical Persian poetry), was at first modeled on the “New Style” (badiʿ) of the early Abbasid poetry, just as the first textbooks on figures of speech were mostly compilations from Arabic treatises with an added set of Persian verses to furnish further examples (on the origins of the Persian ʿelm-e badiʿ and its development see BADIʿ). One should not assume, however, that the creative spirit of Persian letters was of foreign origin. Arab moḥdaṯun poets developed their virtuoso techniques under strong Iranian influence and saturated their verses with both Persian imagery and ways of expression (Danner, pp. 566-94; for a description of the complex interrelation between Arabic and Persian poetry, based on reciprocal influences, see Reinert, pp. 71-105). The “New Style” of decorating poetry with all kinds of rhetorical embellishment originated as much in the Persian as in the Arabic tradition. In a sense the Persians borrowed from the Arabs the very style of poetry that they had earlier helped them create.

New Persian poetry exploited rhetorical figures from its very beginning. The theme of ‘naked verse in need of bedecking’ had become conventional as early as the 10th century, as illustrated by this bayt (line of verse) by Daqiqi (d. 978): “The panegyric came to me naked / It was my splendor and ornaments that garbed it in shirt and mantle” (Rāduyāni, p. 133). But compared to subsequent achievements, the ornamentation in Samanid and even early Ghaznavid poems seems rather modest. During the 11th century we witness an increasing usage not only of rhetorical devices but also of references to these devices in poems, e.g. by Farroḵi (d. 1037) and ʿOnṣori (d. 1039). The first extant treatise on badiʿ figures, Tarjomān al-balāḡa, appeared at the end of the 11th century. These aesthetic tenets (ʿelm-e badiʿ) formed the very core of Persian literary theory, while in neighboring traditions, e.g. Sanskrit, Greek, Latin, and to a certain extent even Arabic, rhetorical figures served only as a complement to other parts of the literary canon.

Towards the end of the 12th century the art of ornamental expression reached its zenith in the panegyric odes (qaṣidas) of the Saljuq court. It was at this time, when the Persian poetic tradition had become well established and the new Arab or ʿErāqi style of poetry was in full bloom, that Rašid-al-Din Waṭwāṭ (d. 1182) presented his treatise Ḥadāʾeq al-seḥr (Gardens of Enchantment). He gave a concise description of the figures popular with Persian poets (63 in all) and selected a set of the best examples for each one. Ḥadāʾeq al-seḥr served as an instructive manual for many generations of poets, shaped the repertory of rhetorical figures, and provided models of usage for centuries to come. It therefore charts the basic directions for the traditional literary work and stylistic tendencies of the classic period. The devices joined and described there under the title of maḥāsen are heterogeneous, having in common only their function as literary ornaments of both poetry and ornate prose (with the exception of some figures unique to versification). Each embellishment introduces some additional organization or parallelism into poetic discourse, which could be phonetic, graphic, lexical, semantic or syntactical. The tripartite division into figures intended to embellish the form (ṣenāʿāt-e lafẓi), the sense (ṣenāʿāt-e maʿnawi), or both the form and the sense of the verse, appeared only in the 15th century (see Atoullo Husayni, Badoyeʿ-uṣ-ṣanoyeʿ, in Tajik). The bayt with its two half-bayts (meṣrāʿ) as a basic structural unit of poetry, or the qarina (context, any part of a sentence) in prose composition provide the framework within which a certain figure is normally displayed (for exceptions, see below).

Phonetic and graphic figures. Word resemblance creates an important resource for embellishment and is presented by Waṭwāt in great detail, partly due to the elaboration of this topic within traditional Arabic grammar. Concepts of Arabic grammar such as lafẓ (verbal expression), ḥarf (letter of the alphabet; prosodic unit corresponding to a grapheme of the Arabic language), ḥarakat (short vowel), number, sequence and shape of the ḥoruf, prosodic pattern (wazn), and meaning (maʿnā) determine the aspects of resemblance.

The central rhetorical figure among those that are concerned with a play on words is tajnis (“making homogeneous,” paronomasia), used in both poetry and prose. The embellishment of the bayt by similar words is subdivided into seven varieties (Waṭwāṭ, pp. 5-12), depending on the kind and degree of similarity. The full (tāmm) tajnis uses homonyms, like the word “ḵaṭā” (meaning both the region of Khotan and a “mistake”) in Waṭwāt’s own line “O, lamp of all the idols of Ḵaṭā, / to be away from your face is a mistake (ḵaṭā)” (p. 6). Other varieties are: tajnis-enāqes (defective or partial tajnis) – words are similar in letters, but different in one short vowel (ḥarakat), like sowār (horse-man) and sewār (bracelet); tajnis-ezāyed (exceeding or superfluous tajnis) – words are similar in letters and short vowels, but one of them ends with an extra letter, like nāl (reed) and nāla (lamentation); tajnis-emorakkab (compound tajnis) – one or both segments are compound, like tā zenda-am (as long as I am alive) and tāzanda-am (I am running); tajnis-e mokarrar (repeated tajnis) – similar words immediately follow one another at the end of the verse or clause and the first of them may have an excess (ziyādat) in the beginning, like basorudorudast (he is with song and music); tajnis-e moṭarraf (tajnis “[different] in ṭaraf,” i.e. last ḥarf) – words are similar in all the letters and short vowels but the last one, like āzār and āzād in del-e karim az āzār āzād bāšad (a generous heart is immune to harm); tajnis-e ḵaṭṭ (tajnis of script) – words share the same character shapes, but differ in pronunciation due to the different diacritical dots above or beyond the characters, cf. “tā” and “bā” in šab-e tārik o rāh-e bārik (a dark night and a straitened path).

Ešteqāq (derivation) can be considered as next of kin to tajnis. It suggests usage of words which are (or appear to be) derived from a common root, like badra (leather purse), badr (full moon) and mobādarat (haste) in the following bayt by Rudaki (d. ca. 940-41): “If the purse (badra) makes you reach for the moon (badr, i.e. the patron) / make haste (mobādaratkon) and do not keep silent!” (Waṭwāṭ, p. 14).

This subtle linguistic approach to the means of assonance is utilized in three more devices. Mokarrar (“repeated”) postulates repetition of the same word in the bayt or two adjoining ones (Waṭwāṭ, p. 86). Maqlub (“overturned” or “reversed”), with its four varieties (pp. 15-17), presents different kinds of palindromes: partial, such as qarib and raqib; full, like mā rā ārām in yā rabb mā rā ārām deh (O Lord, give us peace!); or even covering the whole meṣrāʿ, like ze Naṭanz-am ze Naṭanz-am ze Naṭanz-am ze Naṭanz (I am from Naṭanz...) which could be read from right to left and back (in Arabic script). Radd al-ʿajozʿala’l-ṣadr (“the return of the ending to the beginning”) ranks high among the figures and numbers six varieties (pp. 18-24). The idea is to decorate by tajnis or ešteqāq the marked parts of the bayt, the beginning (ṣadr) and the ending (ʿajoz), or the middle of the first half-bayt (ḥašw) and the ending. The sixth variety (p. 23) is most remarkable in definition: the words which begin and end the bayt seem cognates but in fact are derived from different roots and bear different meaning, like nālam and nāl in the example “I am crying (nālam) day and night in love with that idol, / and thus because of crying (nāla) I have become like a reed (nāl).” The definition here implies the subtle beauty of a false etymology, and the figure needs a perceptive audience attuned to the nuances of the language to be fully appreciated.

Several figures, dealing in some way with similarity between words, provide additional means of cadence and rhyming, not canonized by the established principles of metrics (ʿaruż) and rhyme (qāfiya). Sajʿ (lit. “the cooing of a dove”), considered by Arab scholars as a specific style of rhymed prose composition, is in Persian rhetoric listed among the figures. The first variety, sajʿ-e motawāzi (parallel sajʿ), is recommended by Waṭwāt (p. 14) for prose only, and is designed to decorate the ends of two or more segments (qarina) with the words corresponding in measure (wazn), i.e. prosodic model, number of letters and rhyme letter (rawi), like bāḵta and tāḵta in guy bāḵta wa asb tāḵta (having played the ball and galloped off with his horse). In the second variety, also specific for prose and called sajʿ-e moṭarraf ([equal] in letter ṭaraf) only the rhyme letter is taken into account, cf. Folān rā karam besyār ast o honar bišomār (so-and-so is endowed with great munificence and untold skills). The third variety, sajʿ-e motawāzen (equal in weight) is related to both prose and poetry, with the only parameter of word correspondence being wazn or prosodic pattern. If the bayt or even two adjoining bayts contain symmetrically placed pairs like dalil (a guide) and fasān (a whetstone), ḵazāyen (treasuries) and sarāyer (secrets) or even Raḵš (the name of Rostam’s horse) and tiḡ (a sword or blade), we witness, according to Waṭwāt (p. 15), sajʿ-e motawāzen (also known as mowāzana), which is easy to miss without proper training in traditional Arabic grammar.

Tarṣiʿ (lit. “embellishing with jewels or gold”) consists of dividing a bayt or a sentence into parts by placing in symmetrical order the words which correspond in measure (wazn) and rhyme letter (rawi). In other words, it is the use of parallel (motawāzi) sajʿ in poetry for decorating the bayts with parallelism of cadence and inner rhyme. Rudaki’s line given as an example (Waṭwāṭ, p. 4) “Kas ferestād ba serr andar ʿAyyār marā / ke makon yād ba šeʿr andar besyār marā (ʿAyyar sent someone to me to relay in confidence: do not recall me so frequently in verse)” shows two additional corresponding parts (ferestādmakon yād, ba serr andar – ba šeʿr andar), which complement and enhance the main rhyme and radif in ʿAyyār marā – besyār marā. In Persian treatises this figure usually goes first or second in description that indicates its major role (“adorning with jewels” is one of the most frequently used metaphors for speech decoration in general).

The devices of rhyme enhancement include eʿnāt (“complication”), known also as lozum mā lā yalzam (observance of non-binding rules). In Waṭwāṭ’s rendering (pp. 26-27) this implies the imposition of an extra rhyme letter (ḥarf) before the main one (ḥarf-e rawi) purely for the sake of ornament, since the rhyme would still be complete and perfect without it, like the letter ta in ketāb (a book) and ʿetāb (a reproach). In later tradition the device is treated as any extra obligation imposed on the composition of a poem, most often by repetition of one or two words in all the bayts of the poem. Such poetry was very popular among the court poets of the pre-Mongol era (see illustrations of eʿnāt in Moʿjam by Šams-e Qays, pp. 384-86). Tażmin al-mozdawaj (“the inserting of coupled”) presupposes additional rhyming of two or more adjoining words within the bayt, like zereh and porgereh in the half-bayt of Farroḵi cited by Waṭwāṭ (p. 28): čo ḥalqahā-ye zereh porgereh do zolf-e dotā (the doubled hair-locks are full of knots like mail rings). Ḏu’l-qāfiyatayn (endowed with two rhymes) establishes double rhyme, with penultimate words of the bayts also rhyming with each other. Moraddaf pertains to enriching the rhyme: radif, which consists of one word or more, is repeated after every rhyme of a poem. Waṭwāt notes that this figure is used primarily in Persian poems, while the Arabs do not have it; only the moḥdaṯun poets sometimes used it in order to give their verses a more elaborate structure. On the contrary, most Persian verses are provided with radif, because an elegant radif reveals the poet’s talent and his excellence in speech (Waṭwāṭ, p. 79).

Two figures deal with violation of mono-rhyme and create stanzaic forms by introducing additional rhymes. Mosammaṭ (“beaded”) prescribes a separate rhyme for every five hemistiches of the bayt, while the sixth one carries the basic rhyme of the poem (p. 63). This, popular with Persian poets beginning with Manučehri (d. ca. 1040), is in Waṭwāt’s view the main original type of Mosammaṭ. Another kind (also called mosajjaʿ) suggests the quadripartite division of the bayt, three parts being decorated with sajʿ rhyme, while the fourth carries the main rhyme of the poem. A good example is provided by the famous lines by Moʿezzi (Waṭwāt p. 62): Rabʿ az delam porḵun konam--aṭlāl rā jayḥun konam / Ḵak-e daman golgun konam--az āb-e čašm-e ḵištan (I shall flood the ruined abode with the blood of my heart, I shall turn the ruins into another Oxus, I shall paint red the abandoned soil by the water of my own eyes). Tarjiʿ (lit. “return”) is defined as the division of the qaṣida into several parts with different mono-rhyme in each one. Each part, five to ten bayts, is separated from the next one by a bayt with a different rhyme (Waṭwāt, pp. 85-86). An example is the tarjiʿ qasida by Jamāl-al-Din ʿAbd-al-Razzāq (d. 1192) praising the Prophet, where each of the eleven mono-rhymed stanzas contains eight bayts, the mono-rhyme in each stanza is different from the rest, and the stanzas are separated from each other by inserted bayts with rhyming half-bayts; Šams-e Qays lists the whole of it because of its extraordinary beauty (Šams-e Qays, pp. 401-7).

Graphic devices, manifested only in the written form of the composition, and especially popular among the court officials and scribes (kātebs), provided opportunities for exploiting the possibilities of Arabic script. In raqtā (“black with white spots”) letters with dots and without dots alternate within each word of the bayt or prose passage. In ḵayfā (“with one eye being blue, the other black”) words consisting of letters with dots alternate with words consisting of letters without dots; in moṣaḥḥaf (“wrongly written”) laudation turns into vilification after a change of diacritic signs, in motalawwen (variable) a bayt can be read in two different meters, depending on the phonetic interpretation of the graphemes.

Semantic and syntactical figures. The devices which play on meaning include first of all the figures of figurative expression (majāz) or the tropes: tašbih (simile), esteʿāra (metaphor), tamṯil (analogy), kenāya and talmiḥ (varieties of allusion). Among the most popular ones are also eḡrāq (hyperbole) and tażādd or motażādd (antithesis), both reflecting the modes of expression widely spread in Iranian letters long before Islamic times. Motażādd (also moṭābaqa) is defined as juxtaposition of contrary things in verse or prose (Waṭwāṭ, p. 24). Šams-e Qays adds that things are similar in their opposition (p. 344), i.e. he treats the antithesis as a kind of simile. The elaborate use of contrasting pairs of concepts creates an important part of poetical technique both in court panegyrics and in love-oriented mystical ghazals (where the more subtle possibilities of coincidentia oppositorum can be further explored). Opposite concepts are usually expressed in parallel syntactic constructions; cf. the verse by Masʿud-e Saʿd-e Salmān (d. 1121): Ey sard o garm-e čarḵ kašida / Širin o talḵ-e dahr čašida (O you who have experienced the cold and the hot from heaven, tasted the sweet and the bitter from fate; Waṭwāṭ, p. 25). This is possibly the reason why ʿAṭāʾ-Allāh Ḥosayni qualified antithesis as one of the devices adorning both form and meaning of the verse (Badāyeʿ, p. 161).

A well-developed aspect of speech embellishment is the poetical syntax. Since panegyric demands embellishment par excellence, a great many figures have been designed to reinforce the laudation by syntactic means. Al-madḥ al-mowajjah (a double-barreled praise) is designed to applaud one quality in such a way that it opens the way for another praise, as in Waṭwāṭ’s bayt: “Your sword deals with the life of your enemy / In the manner of your generosity, demolishing a mine of gems” (Waṭwāṭ, p. 36). Taʾkid al-madḥ be-mā yošbeho al-ḏamm (affirming praise through apparent censure) invokes an antithesis, the praise being reinforced by a false note of vilification, mostly by the use of contrastive conjunctions lāken, ammā, etc., cf. the verse by Qamari (fl. second half of the 10th century): “Your friends are forever boasting on account of your glory, nevertheless (lāken) / enemies have to admit that you are peerless” (Waṭwāṭ, p. 38). Estedrāk (emendation) presents another variety of antithesis: the bayt starts with apparent vilification, but develops into praise, like “I do not want my lord to leave a trace in the world. I want my lord to remain in the world forever” (Waṭwāṭ, p. 80). Waṭwāṭ himself advises against the use of the latter device, since the ominous first half of the line heralds an unfavorable augury that may well spoil the patron’s enjoyment of the entire panegyric; in Moʿjam the device is already listed as a defect (p. 295).

Four figures refer to the parts of the panegyric qaṣida where ornament (of any kind) is most desirable. These are the devices designed for special embellishment of the first bayt (ḥosn al-maṭlaʿ “a felicitous beginning”), the bayt wherein the poet moves from the introduction to the praise (ḥosn al-taḵallos “a fine transition”), the final bayt (ḥosn al-maqṭa “a beautiful ending”) and the bayt containing the petition (ḥosn al-ṭalab, an attractive petition). It was well understood that the beauty of the first bayt guaranteed the addressee’s favor, the dexterity of transition attracted his attention to the laudation itself, the final bayt was enunciated last and stayed in memory, and the elegance of the petition ensured the fulfillment of wishes. Especially valued was the art of taḵallos, i.e. of intertwining lyrical or natural motifs with laudatory motifs within one bayt. An acknowledged master of such transitions was ʿOnṣori, as for example in his bayt “If the rose-garden has turned yellow in the autumnal wind, it is as it should be: / It’s only the ruler’s face that should forever stay rubicund.” Waṭwāt quotes it and notes that in this respect ʿOnṣori’s standing among the Persians is comparable to that of Motanabbi amidst the Arabs (p. 32).

A number of artifices are connected with a certain syntactical shaping of the poetical idea and modes of its presentation. Eʿterāż al-kalām qabl al-tamām (interruption of speech before its completion) provides some pleonastic interpolation (ḥašw) for the idea of the bayt; it could be of three kinds, ugly (qabiḥ), neutral (motawasseṭ) and elegant (maliḥ). In this last case the elegance of the insertion justifies the violation of the semantic and syntactic integrity of the bayt, as in ke borranda bādā (may it be forever incisive!) in Waṭwāṭ’s own example “The images of his sword – may it be forever incisive! / – Have lodged in the souls of his foes” (Waṭwāṭ, p. 54). An almost similar device is eltefāt (“turning towards another”) in its Persian interpretation (Waṭwāṭ, p. 38): the meaning is expressed completely and then something relevant is added (proverb, prayer, etc.), like the second hemistich in the bayt of Monjik (d. 981 or 991): “The arrow of our separation from you wounded the heart. / O patience, you make a fine coat of mail against separation from idols!” (p. 39). Soʾāl o javāb (question and answer) defines constructing the bayt or bayts as questions and answers. This form may be traced to the Iranian folklore tradition. Waṭwāṭ notes that it is highly regarded among the Persians (p. 59), and prominent poets like Farroḵi and Moʿezzi (d. between 1124 and 1127) composed whole qasidas in this manner (for the translation of one dialogic nasib by ʿOnṣori see Browne, II, p. 121-23).

Jamʿ wa tafriq wa taqsim (combining, differentiating and apportioning) prescribes to shape the poetic idea by establishing a feature which joins (jāmeʿ) the described objects, or by describing differences between them, or by enumerating first the objects and next their attributes. It is subdivided into six varieties, jamʿ, tafriq, and taqsim proper, jamʿ with tafriq, jamʿ with taqsim and jamʿ with tafriq and taqsim, and involves different kinds of comparison and antithesis. The last and most difficult variety is illustrated in Waṭwāṭ (p. 77) with the do-bayti (quatrain) “What chained you, chained your slave as well, / With a chain not manifest but invisible. Your chain is made of iron and mine of sorrow, / Your chain is on your feet and your slave’s on his heart (soul).” Thus, in Waṭwāt’s words, the poet first has brought together (jamʿ karda) his beloved and himself, for they are both in chains, then he has distinguished (tafriq karda) between the manifest and the invisible, and in the second bayt he has distributed (taqsim karda) the attributes by specifying the material (iron, sorrow) and place (feet, soul) of each chain. This complex device based on simile allows the poet to describe the poetic persona and his/her object of love or praise simultaneously in terms of similarity and opposition; it gained considerable popularity already during the Ghaznavid period, cf. the praise of ʿOnṣori’s craft in composing whole qasidas in that manner (Rāduyāni, p. 68), as well as Waṭwāt’s remark that Persians are masters of using taqsim in every bayt of the qasida (Waṭwāt, p. 76; for some fine specimens by Farroḵi, Manučehri and Adib Ṣāber (d. between 1144 and 1148) with commentary see Bertels, 1960, pp. 338, 368, 511).

Closely associated with taqsim is tafsir-e jali wa ḵafi (explicit and implicit commentary): a poet mentions several words which seem obscure and then repeats them with proper commentary (mostly using parallel syntactical constructions); as in ʿOnṣori’s lines celebrating the ruler’s might: “He either puts in chains (bandad) or forces open (gošāyad, “opens”), or takes and gives / And as long as the world lasts, this will serve as the king’s lasting memento: What he takes is the country, what he gives is bounty / What he enchains is the (foot of) the enemy, what he conquers is the fort” (Waṭwāṭ, p. 78).

Siyāqat al-aʿdād (enumerating) and tansiq al-ṣefāt (arranging attributes) relate to models of description. In the first case different things belonging to somebody or something are enumerated. In the second, different attributes describing some person or a thing are enumerated. It is strongly recommended to combine enumeration with some figures of word semblance, like tajnis, or of semantic opposition, like antithesis (Waṭwāt, p. 50).

Morāʿāt al-naẓir (“the observance of the similarity”) which is generally known under the name of tanāsob (congruity), pertains to bringing together several homogeneous (az jens-e yek degar) things within the frame of a bayt, like moon and sun, arrow and bow, lips and eyes, rose and tulip. Waṭwāt remarks that few Arabic or Persian poems lack this device, but the degree of elegance varies (p. 35). The congruity of tropes as an aesthetic principle has been described by Šaraf-al-Din Rāmi (pp. 54-58) as a stringent regulation for figurative speech.

The devices specifically designed to play on the conventionality of themes and motifs are also included into the set of figures. They give the poet the possibility to de-familiarize (the device of making it strange, “priëm ostraneniija,” using the terminology of the Russian formalists) the commonplace imagery, and transform it into a new and freshly invented entity. Tajāhol al-ʿāref (the feigned ignorance of one who already knows) prescribes to present a conventional idea as still potentially arguable, transforming it into a (rhetorical) dilemma, as in “O Lord, is he the ruler of the world or the world itself?” as opposed to the topos “the ruler comprises the whole world” (Waṭwāṭ, p. 58). Taʿajjob (wondering) is also a form of rhetorical question, which expresses wonder (on wonder as the pleasurable quality of poetry, see Yarshater, 1988, pp. 268-69 and n. 70), as in the line by Waṭwāt: “Can’t you tell why I have water forever welled up in my eyes?/ For you have a well (a dimple) - I know too well - on your chin!” (Waṭwāṭ, p. 84). The fact that the water (tears) covers the eyes (common attribute of lovers) is described as causing wonder, because water comes from the well (čāh), and it belongs to the beloved (the dimple on the chin is regarded as a beautiful attribute in Persian love poetry). In ḥosn al-taʿlil, the fantastic conjunction of conventional ideas is not only implied but also explicitly presented within the bayt. Though ḥosn al-taʿlil ranked as an embellishment, the elaboration of this figure within the ʿErāqi style of poetry proves it to be more of a way of combining images than a mere ornament.

Poetry could be embellished by several types of quotation. Tażmin (“including one thing into another”) consists in borrowing a quotation from other poet’s verses; the advisable length is from half to two bayts and the passage should be very well-known or marked in some way to avoid the charge of plagiarism (Waṭwāṭ, p. 72; on an alternative handling of tażmin as syntactical dependence of two adjoined bayts, see Šams-al-Din Rāzi, p. 290). Ersāl al-maṯal (inclusion of a proverb) and ersāl al-maṯalayn (inclusion of two proverbs) provide the ornamentation of the verse with gnomes, maxims and didactic sentences. The passion for didacticism could be traced in all the genres of Iranian literature, so both devices were very much in use (see ANDARZ; EPIGRAM); some poets even veiled their own arguments under the guise of proverbial form (on ersāl in ʿOnṣori’s Divān see Bertels, 1988, pp. 143-46). Closely related to this is the device al-kalām al-jāmeʿ (comprehensive discourse), whereby gnomic observations, edifications and complaints, are all brought into the poetry. This device is characteristic of the ḥabsiyyāt (prison poems) genre, as in the prison poems of Masʿud-e Saʿd-e Salmān which, in Waṭwāt’s opinion, surpassed all others in this style (p. 82).

Two devices reflect the connections and interplay with Arabic poetry. These are molammaʿ (“speckled”), a macaronic poem alternating Arabic and Persian bayts or half-bayts, and tarjama (translation), rendering Arabic bayts into Persian or vice versa; Rāduyāni also includes chapters (pp. 118-27) on such figures as the versification of Ḥadith, Arabic proverbs and wise sayings (tarjamat al-aḵbār wa-’l-amṯāl wa’l-ḥekma), on parallels between Persian parables (afsāna) and Qurʾānic verses and on rendering the meaning of Qurʾānic verses by Persian verses, but these devices are apparently not employed in later tradition.

Numerous devices offer opportunities to extract a hidden meaning from a bayt or a poem. Ihām (amphibology, double-entendre) privileges a second and less obvious meaning embedded in a line in lieu of the primary and obvious sense; loḡ¯z, also loḡ¯az (a puzzle) is a poem constructed as a series of questions and called čistān (“what is that?”) in Persian, and it is closely related to moʿammā (a logogriph), which need not be in the format of a question. The name of the beloved or other appellation could be encrypted by some sophisticated comparison, or a certain way to find the answer could be implied, e.g. recombination and transposition of the graphic characters and diacritic signs (Waṭwāṭ, p. 70). Later, the popularity of this genre increased so that Šams-e Qays in his Moʿjam enrolled moʿammā into his categories (ajnās) of poetry, and in the 15th century special treatises were already devoted to various types of charades (Seyed-Gohrab, 2001, pp. 15-36). The mowaššaḥ (“girdled”) is an acrostic poem, i.e. with some letters, words or parts of the poem revealing an encrypted text when combined together. Mādda-ye tāriḵ (chronogram) contains the date of an important event by using the numerical values of the abjad system, and has been listed as a figure of speech in Badāyeʿ (Ḥoṣayni pp. 159-60; on chronographic virtuosity and “feats of formal acrobatics” of the Timurid poets, see Yarshater, 1988, p. 270). ʿAṭā-Allāh Ḥoṣayni also describes (p. 159) sehr-e ḥalāl (licit magic; often used in a wider sense as a metaphor for poetry tout court), where one can extract several different meanings of the verse by joining the “magic” word group with any of the preceding or following words of the bayt as in the line “Ay šāh ba jān rasida kāram digar / Az ḡoṣṣa o ḡamfegān ke jān ḵᵛāham dād”; the phrase az ḡoṣṣa o ḡam (from grief and sorrow) could be syntactically combined with ba jān rasida kāram, and the meaning of the bayt would be “O king, I am on the verge of death because of grief and sorrow, O woe that I am about to die;” it could be combined also with jān ḵāᵛham dād, and the verse would mean “O king, I am on the verge of death, what a pity that I am to die of grief and sorrow.” The examples given for seḥr-e ḥalāl in another 15th century treatise, Badāyeʿ al-afkārfi ṣanāyeʿ al-ašʿār, are discussed by its editor, Mir Jalāl-al-Din Kazzāzi in his commentary (Kāšefi, pp. 306-8; see also Simidchieva, p. 528). This narrower usage of sehr-e ḥalāl had been noted in earlier treatises, however in Badāyeʿ al-ṣanāyeʿ (pp. 159-60), taʿajjob, seḥr and tāʿriḵ create a separate and distinguished group of figures of meaning (ṣanʿathā-ye maʿnawiya) which had not been considered as ornaments by Arab masters of eloquence and had become considered as such only by Persians.

The problem of stylistic unity and arrangement of the poem as a whole, and not just within a single bayt, was reflected in the formulation of tafwif (“white-striped cloth”). This figure, previously absent from Persian badiʿ books, heads the list of figures in Moʿjam (pp. 329-35). It recommends maintaining a uniform style and manner throughout a qaṣida and requires a harmonious combination of all poetic elements (meter, rhyme, wording, ideas, poetic figures) both horizontally, i.e. within the bayt, and vertically, throughout the whole qaṣida to achieve a symmetrical structure. The examples for this figure by Anwari (d. ca. 1169), Rašid-al-Din Waṭwāt, Ẓahir Faryābi (d. ca.1201) and Rażi-al-Din Nišāpuri (d. c. 1201-02) amount to 86 bayts in all, since tafwif can only be displayed and appreciated in poems of some length. In Šams’ opinion (p. 444) the chapter on tafwif in his book contains truly skilful and natural (maṭbuʿ) poetry.

It is in terms of natural (maṭbuʿ), artificial (maṣnuʿ) and mannered (motakallaf) poetry that medieval Persian poets and critics used to express their attitude towards ornamentation of speech. Tarjomān contains the early formulation: while studying the figures of speech one should not forget that the easier (sahltar) and simpler (bitakalloftar) the speech (soḵan), the better (p. 111). However, the great masters like ʿOnṣori are capable of using a great variety of figures in their qaṣidas without any loss to their ṭabʿ “naturalness, talent” (p. 69). Šams-e Qays takes into account the stylistic achievements of ʿErāqi style. In his opinion all poetic devices and figures of embellishment belong to varieties of the artificial (motakallafāt) in poetry, attainable only through careful attention and deep contemplation (Šams-e Qays, p. 432). He proposes to label as maṭbuʿ the verses where takallofāt are employed with elegance and a feeling for measure. He probably means that in such a case, due to the poet’s craftsmanship, artificial poetic devices are perceived as natural language creations. In books on badiʿ such an extreme limit of craftsmanship was called sahl-e momtaneʿ “of inimitable simplicity” (e.g. Badāyeʿal-ṣanāyeʿ, p. 204; cf. sahl wa momtaneʿ in Waṭwāṭ, p. 87), and authors of poetic anthologies awarded this epithet to the style of many great poets. In a sense this stylistic characteristics is analogous to the category of eʿjāz (inimitability of the Qurʾān) in the Arabic tradition, being, however, unlike the latter, achievable in principle (on the striking rhetorical displays aimed at “imitation of the inimitable” see Subtelny, 1986, pp. 56-79).

The number of figures increased considerably with time. The complex and combined rhetorical devices created an important part of the intrinsic difficulties for which the Indian style is famous (on the Safavid poets’ ingenuity in crossing rhetorical figures, see Yarshater, 1955, p. 190ff.). Ḥosayn Wāʿeẓ Kāšefi (d. circ. 1504), author of the extremely ornamental Anwār-e Sohayli, already describes more than 200 devices and varieties in his Badāyeʿ al-afkār fi šanāyeʿ al-ašʿār, introducing numerous subtypes for already known figures, without adding new ones (for details, see Simidchieva).

Rhetorical ornament in Persian poetry serves, as Julie Scott Meisami puts it, to reveal the “hidden truth” of the matter and to increase the effectiveness of its presentation (Meisami, p. 309). Figures of speech are not only the conceptual component of this poetry but also its eventful domain. The true poetic event happens more often than not in the realms of conventional language, when some elaborate combination of rhetorical devices has been dexterously applied; it could only be appreciated by the genuine connoisseurs capable of perceiving the art. This kind of poetry leaves but a small chance to the translator (hence the lamentations on the richness and figurativeness of Persian style as excessive to the Western mind and taste). Traditional figures of Persian rhetoric, when being analyzed in their stylistic function and expressive potential and not just enumerated, serve as useful tools for describing the poet’s style and imagery (cf. the analysis of poetic structure in Meisami, 2003; the description of ʿOnṣori’s poetic technique in Bertels, 1988, pp. 129-55; for Neẓāmi’s see Seyed-Gohrab, 2003, pp. 30-40).


Evgenĭ E. Bertel’s, “Khakim ʿUnsuri iz Balkha” (Ḥakim ʿOnṣori of Balḵ), in idem, Izbranyye trudy V: Istoriya literatury I kul’tury Irana, ed. G. Ju. Aliev and N. Prigarina, Moscow, 1988, pp. 8-201.

Idem, Istoriyapersidsko-tadzhikskoĭ literatury, Moscow, 1960.

S. A. Bonebakker, “Luzūm mā lā yalzam,” in EI² V, pp. 839-41.

Idem, “Tawriya,” in EI² X, pp. 395-96.

Edward G. Browne, Literary History of Persia, 4 vols., Cambridge, 1929-30, II, p. 17 ff.

Victor Danner, “Arabic Literature in Iran: The Umayyad Age in Iran,” in Cambridge History of Iran IV, pp. 566-94.

E. J. W. Gibb, A History of Ottoman Poetry, 6 vols., London, repr. 1958-63.

Jalāl-al-Din Homāʾi, Fonun-e balāḡat wa ṣenāʿāt-e adabi, 2 vols. in 1, Tehran, 1975.

Atoullo Husayni (ʿAṭāʾ-Allāh Ḥosayni), Badoyeʿ-uṣ-ṣanoyeʿ (Badāyeʿ al-ṣanāyeʿ), ed. Rahim Musulmankulof, Dushanbe, 1974.

ʿOnṣor-al-Maʿāli Kaykāvus b. Eskandar, Qābus-nāma, ed. Ḡolām-Ḥosayn Yusofi, repr. Tehran, 1999.

Kamāl-al-Din Ḥosayn Wāʿeẓ Kāšefi, Badāyeʿ al-afkārfi ṣanāyeʿ al-ašʿār, ed. Mir Jalāl-al-Din Kazzāzi, Tehran, 1990.

Mir Jalāl-al-Din Kazzāzi, Zibā-šenāsi-e soḵan-e pārsi III: Badiʿ, Tehran, 1994.

Julie Scott Meisami, Medieval Persian Court Poetry, Princeton, 1987.

Idem, Structure and Meaning in Medieval Arabic and Persian Poetry, London, 2003.

R. Musulmankulof, Persidsko-tadjikskaya klassicheskaya poetica X-XV vv, Moscow, 1989 (detailed comparative analysis of the figures’ definitions in 11 treatises on badiʿ from the classical period).

Moḥammad b. ʿOmar Rāduyāni, Tarjomān al-balāḡa, ed. Ahmed Ateş, Istanbul, 1949.

Šaraf-al-Din Ḥasan Rāmi, Anis al-ʿoššāq, ed. ʿAbbās Eqbāl, Tehran, 1946.

Benedikt Reinert, “Probleme der vormongolischen arabisch-persischen Poesiegemeinschaft und ihr Reflex in der Poetik,” in Gustave E. von Grunebaum, ed., Arabic Poetry: Theory and Development, Wiesbaden, 1973, pp. 71-105.

Šams-al-Din Moḥammad b. Qays Rāzi, al-Moʿjam fi maʿāyir ašʿār al-ʿAjam, ed. Moḥammad Qazvini, rev. Moḥammad-Taqi Modarres Rażawi, Tehran, 1959; tr. Natalia Chalisova as Svod pravil persidskoy poezii, Moscow, 1997, commentary pp. 346-428.

Gregor Schoeler, “Muwashshaḥ,” in EI² VII, pp. 809-12.

A. A. Seyed-Gohrab, “The Art of Riddling in Classical Persian Poetry,” Edebiyāt 12, 2001, pp. 15-36.

Idem, Layli and Majnun: Love, Madness, and Mystic Longing in Nizami’s Epic Romance, Leiden, 2003.

Marta Simidchieva, “Imitation and Innovation in Timurid Poetics: Kashifi’s Badāyiʿ al-afkār and Its Predecessors, al-Muʿjam and Ḥadāʾiq al-siḥr,” Iranian Studies 36/4, 2003 pp. 509-30.

Maria E. Subtelny, “A Taste for Intricate: The Persian Poetry of the Late Timurid Period,” ZDMG 136, 1986, pp. 56-79.

Rašid-ad-Din Waṭwāṭ, Ḥadāʾeq al-seḥr fi daqāʾeq al-šeʿr, ed. ʿAbbās Eqbāl, Tehran, 1929 (repub. in his Divān, pp. 621-707); ed. and tr. Natalia Chalisova as Sady volshebstva v tonkostyah poezii, Moscow, 1985, commentary pp. 172- 202.

Idem, Divān, ed. Saʿid Nafisi, Tehran, 1960.

Ehsan Yarshater, “Safavid Literature: Progress or Decline,” Iranian Studies 7/1-2, 1974, pp. 217-79.

Idem, “The Development of Iranian Literatures,” in idem, ed., Persian Literature, Columbia Lectures on Iranian Studies 3, Albany, 1988, pp. 3-37.

Idem, “The Indian Style: Progress or Decline,” in ibid., pp. 249-88.

Idem, Šeʿr-e fārsi dar ʿahd-e Šāhrok: āḡāz-e enḥeṭāṭ dar šeʿr-e fārsi, 2nd. ed., Tehran, 2004.

July 20, 2009

(Natalia Chalisova)

Originally Published: July 20, 2009

Last Updated: July 20, 2009