Sanskrit prosody

From Wikipedia, the free encyclopedia
Jump to: navigation, search
See Vedic meter for meter in Vedic Sanskrit.

Sanskrit prosody is the study of metre in Sanskrit poetry. Standard traditional works on meter are Pingala's Chandaḥśāstra and Kedāra's Vṛttaratnākara. The most exhaustive compilations, such as the modern ones by Patwardhan and Velankar contain over 600 meters. This is a substantially larger repertoire than in any other metrical tradition.[1]


The metres of classical Sanskrit poetry are divided into three kinds.[2]

  1. Syllabic verse (akṣaravṛtta): meters depend on the number of syllables in a verse, with relative freedom in the distribution of light and heavy syllables. This style is derived from older Vedic forms, and found in the great epics, the Mahabharata and the Ramayana.
  2. Syllabo-quantitative verse (varṇavṛtta): meters depend on syllable count, but the light-heavy patterns are fixed.
  3. Quantitative verse (mātrāvṛtta): meters depend on duration, where each verse-line has a fixed number of morae, usually grouped in sets of four.

Elements of prosody[edit]

Light and heavy syllables[edit]

In most of Sanskrit poetry the primary determinant of a meter is the number of syllables in a unit of verse, called the pāda ("foot" or "quarter", not to be confused with the "foot" of Western prosody). Meters of the same length are distinguished by the pattern of laghu ("light") and guru ("heavy") syllables in the pāda.

The rules distinguishing laghu and guru syllables are the same as are specified in Vedic texts such as the Pratisakhyas. They can be summarized as:[3][4]

  1. A syllable is laghu only if its vowel is hrasva ("short") and followed by at most one consonant before another vowel is encountered.
  2. A syllable with an anusvara ('ṃ') or a visarga ('ḥ') is always guru.
  3. All other syllables are guru, either because the vowel is dīrgha ("long"), or because the hrasva vowel is followed by a consonant cluster.
  4. The hrasva vowels are the short monophthongs: 'a', 'i', 'u', 'ṛ' and 'ḷ'
  5. All other vowels are dīrgha: 'ā', 'ī', 'ū', 'ṝ', 'e', 'ai', 'o' and 'au'. (Note that, morphologically, the last four vowels are actually the diphthongs 'ai', 'āi', 'au' and 'āu', as the rules of sandhi in Sanskrit make clear. So, while an original 'ai', for example, had been shortened to an 'e' sound in practice, it was still to be treated as long metrically. The original short 'e' and short 'o' sounds had already been assimilated into short 'a' in the Proto-Indo-Iranian period of the language.[5])
  6. Gangadasa Pandita states that the last syllable in each pāda may be considered guru: सानुस्वारश्च दीर्घश्च विसर्गी च गुरुर्भवेत् । वर्णः संयोगपूर्वश्च तथा पादान्तगोऽपि वा ॥. But a guru at the end of a pāda is never counted as laghu.

For measurement by mātrā (morae) , laghu syllables count as one unit, and guru syllables as two units.[6]

In Prakrit[edit]

In Prakrits (and their modern descendents like Awadhi), prosody has different rules than Sanskrit. A Laghu before a Samyukta in a different word is not considered Guru in Prakrit. Neither are laghu-s before a few conjuncts like प्र, ह्र, ब्र and क्र.


Gaṇa (Sanskrit, "group") is the technical term for the pattern of light and heavy syllables in a sequence of three. It is used in treatises on Sanskrit prosody to describe meters, according to a method first propounded in Pingala's chandaḥśāstra.

Pingala's method described any meter as a sequence of gaṇas, or triplets of syllables, plus the excess, if any, as single units. There being eight possible patterns of light and heavy syllables in a sequence of three, this scheme called for ten descriptive elements in all. With each of these ten, Pingala associated a letter, allowing the meter to be described compactly as an acronym. His encoding scheme was as follows[7]

  • The units:[8]
  • l: a "light" syllable (L), called laghu
  • g: a "heavy" syllable (H), called guru
  • The gaṇas:[9]
  • m : H-H-H, called ma-gaṇa
  • y : L-H-H, called ya-gaṇa
  • r : H-L-H, called ra-gaṇa
  • s : L-L-H, called sa-gaṇa
  • t : H-H-L, called ta-gaṇa
  • j : L-H-L, called ja-gaṇa
  • bh: H-L-L, called bha-gaṇa
  • n : L-L-L, called na-gaṇa

Pingala's order of the gaṇas, viz. m-y-r-s-t-j-bh-n, corresponds to a standard enumeration in binary, when the three syllables in each gaṇa are read right-to-left with H=0 and L=1.

An example[edit]

The definition of the meter Vasantatilakā given by Kedāra in his Vṛttaratnākara is

uktā vasantatilakā tabhajā jagau gaḥ

This definition is itself an example of the meter, but the list of gaṇas contained in it can be decoded as:[10]

tabhajā jagau gaḥ = t bh j j g g = H-H-L-H-L-L-L-H-L-L-H-L-H-H

A mnemonic[edit]

The word yamātārājabhānasalagāḥ (or yamātārājabhānasalagaṃ), invented by medieval commentators, is a mnemonic for Pingala's gaṇas, using the vowels "a" and "ā" for light and heavy syllables respectively with the letters of his scheme. In the form without a grammatical ending, yamātārājabhānasalagā is self-descriptive, where the structure of each gaṇa is shown by its own syllable and the two following it:[11]

  • ya-gaṇa: ya-mā-tā = L-H-H
  • ma-gaṇa: mā-tā-rā = H-H-H
  • ta-gaṇa: tā-rā-ja = H-H-L
  • ra-gaṇa: rā-ja-bhā = H-L-H
  • ja-gaṇa: ja-bhā-na = L-H-L
  • bha-gaṇa: bhā-na-sa = H-L-L
  • na-gaṇa: na-sa-la = L-L-L
  • sa-gaṇa: sa-la-gā = L-L-H

The mnemonic also encodes the light "la" and heavy "gā" unit syllables of the full scheme.

The truncated version obtained by dropping the last two syllables, viz. yamātārājabhānasa, can be read cyclically (i.e., wrapping around to the front). It is an example of a De Bruijn sequence.[12]


The gaṇas are not the same as prosodic feet in Greek or Latin poetry, although there is a correspondence (m-y-r-s-t-j-bh-n = molossus, bacchius, cretic, anapest, antibacchius, amphibrach, dactyl, choreus). The difference is that the gaṇas are analytic devices only, and do not indicate internal structure as "feet" do. For instance, a phalaecian verse consisting of a spondee, a dactyl and three trochees would be analysed as m-s-j-g-l (i.e. a molossus, an anapest, an amphibrach and a trochee); similarly a sapphic verse as r-t-j-g-l (cretic, antibacchius, amphibrach and trochee).[13]

Matra gaNa-s[edit]

The standard unit of grouping, analogous to the "foot" of Western prosody, is four morae (four laghus, two gurus, or a guru and two laghus).


Most of classical Sanskrit poetry is of the varṇavṛtta type, also called akṣarachandas. Stanzas are quatrains of four pādas (verses), with the metrical structure of each pāda completely specified. In some cases, pairs of pādas may be scanned together as the hemistichs of a couplet.[14] It is then normal for the pādas comprising a pair to have different structures, to complement each other aesthetically. Otherwise the four pādas of a stanza will have the same structure.

Examples of Akṣarachandas[edit]

Indravajrā (11 Syllables)

Definition - syād indravajrā yadi tau jagau gaḥ

Translation - The Indravajrā meter contains ta ta ja ga ga. (The caesura (yati) is after the fifth syllable.)

Meticral Scheme - - u - - | u u - u - -

Śikhariṇī (17 Syllables)

Definition - rasai rudraiś chinnā yamanasabhalā gaḥ śikhariṇī

Translation - The Śikhariṇī meter, divided by the 6 flavors and the 11 Rudras, contains ya ma na sa bha la and ga. (The caesura (yati) is after the sixth syllable.)

Metrical Scheme u - - - - - | u u u u u - - u u u -

Śārdūlavikrīḍita (19 Syllables)

Definition - sūryāśvair masajāstataḥ saguruvaḥ śārdūlavikrīḍitam or sūryāśvair yadi maḥ sajau satatagāḥ śārdūlavikrīḍitam

Translation - The Śārdulavikrīḍita meter contains ma sa ja sa ta ta plus on heavy syllable. The caesura (yati) divides the verse into 12 (sūrya) and 7 (aśva) syllables, i.e. it occurs after the twelfth syllable.

Meticral Scheme - - - u u - u - u u u - | - - u - - u -

Epic poetry[edit]

While the Mahabharata has various types of versification, an overwhelming proportion of the stanzas (all but about 0.2%) are akṣaravṛtta (free syllabic). Within this majority, 95% are shlokas of the anustubh type and the rest are tristubhs.[15]

Moraic poetry[edit]

See also: Mātrika metre

Some Sanskrit forms are measured in units of time—one for a short syllable, two for a long one—called mātrās, similar to the morae of Greek verse. The cover term āryā can be used to refer to the Sanskrit moraic forms generally, though it also names a particular variant.

  • mātrāchanda:[16]
  1. puṣpitāgrā
  2. aparavaktra
  3. vaitālīya
  4. mātrāsamaka
  1. āryā
  2. āryāgīti
  3. upagīti

Problems of enumeration[edit]

In the Indian tradition of prosody, the names of the Vedic metres are used as collective names for all metres with a certain number of syllables per line.[18] Thus for instance,

  • the anuṣṭubh class refers to metres with 8 syllables per line (the śloka metre being the most popular; others include vidyunmālā)
  • the triṣṭubh class refers to metres with 11 syllables per line (this includes indravajrā, upendravajrā, rathoddhatā, śālinī, and other metres)
  • the jagati class refers to metres with 12 syllables per line, etc.

As a theoretical exercise, the Indian writers on metrics became interested in problems related to enumerating all possible lines (patterns of laghu and guru), the algorithms for which became known as pratyaya. For a given class (length), the six pratyaya were:[19]

  • prastāra, the "table of arrangement": a procedure for enumerating (arranging in a table) all metres of the given length,
  • naṣṭa: a procedure for finding a metre given its position in the table (without constructing the whole table),
  • uddiṣṭa: a procedure for finding the position in the table of a given metre (without constructing the whole table),
  • laghukriyā or lagakriyā: calculation of the number of metres in the table containing a given number of laghu (or guru) syllables,
  • saṃkhyā: calculation of the total number of metres in the table,
  • adhvan: calculation of the space needed to write down the prastāra table of a given class (length).

Some authors also considered, for a given metre, (A) the number of guru syllables, (B) the number of laghu syllables, (C) the total number of syllables, and (D) the total number of mātras, giving expressions for each of these in terms of any two of the other three. (The basic relations being that C=A+B and D=2A+B.)[20]


  1. ^ Deo, p. 3
  2. ^ Deo, p.5
  3. ^ Coulson, p.21
  4. ^ Muller & Macdonell, Appendix II
  5. ^ Coulson, p.6
  6. ^ Muller and Macdonell, loc.cit.
  7. ^ Pingala, chandaḥśāstra, 1.1-10
  8. ^ Pingala CS 1.9-10, in order
  9. ^ Pingala CS, 1.1-8, in order
  10. ^ Coulson, p.253
  11. ^ Coulson, p.253ff
  12. ^ Stein, Sherman K. (1963), "Yamátárájabhánasalagám", The Man-made Universe: An Introduction to the Spirit of Mathematics, pp. 110–118 . Reprinted in Wardhaugh, Benjamin, ed. (2012), A Wealth of Numbers: An Anthology of 500 Years of Popular Mathematics Writing, Princeton Univ. Press, pp. 139–144.
  13. ^ Colebrooke, p.64
  14. ^ Hopkins, p.194. (This is typical for the shloka).
  15. ^ Hopkins, p.192
  16. ^ Hopkins, p.193
  17. ^ Hopkins, loc. cit.
  18. ^ ‌Hahn, p. 3.
  19. ^ Hahn, p. 4
  20. ^ Hahn, pp. 15–18

See also[edit]


External links[edit]