(Bengali setār; Marathi satār).
Large, fretted long-necked lute. It is a prominent instrument of the classical music of the northern and central regions of South Asia. Though played by some itinerant rural performers (in Rajasthan) and in modern film and radio orchestras, the sitār is found mainly in chamber music as formerly practised at the Muslim and Hindu courts and now on the public concert stage and through urban media.
2. Modern structure and tuning.
The name sitār is an Urdu transcription of sihtār (‘three-stringed’) from Persian, the court language of North India from the 13th century to the 19th. It did not become standard in India until the instrument began to reach its present form in the 18th century, and we must look to other Perso-Turkic names for long-necked lutes – tanbūr, tanbūrah, to which sihtār may first have been an adjective – for early forms of the sitār, as well as of the tambūrā, which is of fundamentally similar construction.
The history of the Hindustani sitār begins with the Muslim Delhi Sultanate (1192–1526). The immigrant Turks and Persians brought their music and instruments with them. In the works of the Indo-Turkish court poet Amir Khusrau (1253–1325) the tanbūr is recorded together with other Islamic instruments. The general nature of the tanbūr can safely be inferred from surviving Perso-Turkish long-necked lutes. These are slender instruments with an ovoid or pear-shaped wooden shell, frontal or right unilateral pegs (or both) inserted directly into the neck (without a separate pegbox), wooden soundtable and gut or fibre strings. They may have included both the fretted forms (with simple tied gut, like the modern western Central Asian tanbūr, dutār and setār) and unfretted (like the Central Asian dambura). A Muslim tradition in South Asia credits the invention of the sitār (and often also of the tablā and the song form khayāl) to Amir Khusrau himself, although these names are not mentioned in his works. The tanbūr continues to be mentioned at the courts throughout this period: Sultan Sikandar Lodi, an Afghan (reigned 1489–1517), had ‘four boy slaves, skilled in chang, rabāb, tanbūr and bīn’.
The early Mughal emperors (1526–1707) and their tribesmen not only brought fresh Central Asian influences to India but were also, especially Akbar, Jahangir and Shah Jahan (who ruled consecutively 1556–1658), great patrons of the arts. The long-necked lute is recorded as tanbūrah at this time, and in Abul Fazl’s contemporary account (c1590) of Akbar’s chief court musicians four players (more than for any other instrument) are named, all Muslim.
Among the tanbūrāh players at the court of Shah Jahan were one Shauqi (‘well-versed in Persian and Hindi melodies’) and his pupil Tara Chand, titled Kalavant (a high Hindu musician caste); clearly the instrument was being adopted by Hindu musicians and for Hindu music. Links are also found with the Uzbek dutār, in the technique used for plucking and in the early three-string tuning, which became well established in the following century when the name sitār became current. Willard (A Treatise on the Music of Hindustan, Calcutta, 1834) gives the three-string tuning as two brass strings tuned in unison to the tonic (kharaj), and the first, of steel, to the 4th; ma–sa–sa. This remains the basis of the later more complex tunings which can be seen as expanded dutār tuning.
The Hindustani court sitār took on the outline of its modern form as a solo instrument of art music during the late Mughal empire (1707–1858). Changes included the widening and thickening of the neck, now always straight, not tapering; the use of gourd for the shell instead of carvel-built ribs; the adoption of heavy metal frets as well as of a rigid nut and string-guider of bone etc. (frequently also simple bindings on non-Indian types); and the fitting of heavy metal strings. The dutār pattern of two frontal pegs was retained, leaving space to the left of the main string (which runs along the frets almost centrally) for large-scale string deflection for mind (portamento).
The name sitār became established early in this period, though we hear of ‘fretted and unfretted tambūrā’ up to about 1800, to distinguish the sitār from the fretless drone tambūrā. It was during the 18th century, however, that the five-string sitār (whose pegs are retained in the modern instrument, placed above the nut) came into use. Several tunings are given in later sources, with two strings (variously drone, melody or cross-plucking) added to the earlier three. With the addition of the cikārī they form the basis of the modern seven-string tuning. 17 frets were common, with variant degrees provided only for the lower and middle 4ths and lower 7th of the first string.
During the late Mughal period the two main bāj (repertories) of gat (composition) were established. The Delhi bāj is attributed to the 18th-century court musician Masit Khan. The descendants of Masit Khan were associated with a large sitār with two extra full-gourd resonators, an instrument of the hybrid bīn-sitār type, and they claimed descent from Tansen (Akbar’s chief court musician). The second is that of the ‘eastern’ (pūrab), or razākhānī, bāj linked to Ghulam Raza Khan (early 19th century) of Lucknow, whose nawabs had by then succeeded the Mughals as the main patrons of Indo-Muslim culture. (See India, §III, 6(i).)
Other sitār types have been found on the periphery of northern and central South Asia, whose distribution and related features suggest a common development. These include the Karnatak sitār, the Kashmiri setār, the Afghan tanbur and perhaps the Gujarati sittarae, which share the pattern of three frontal pegs and a main string in a double course, and are all accompanying or band-sitārs.
Several types of sitār can still be found in manufacture, but the most common in Hindustani concert music is fairly uniform, with two main models, single-gourd and double-gourd (fig.1); other types include the sūrbahār (see below) and the rarely found kachvā sītār.
Tarafdār sitār means ‘sitār with sympathetic strings’, but this type is often simply called sitār, this feature having become standard. It is made of wood and a bulging gourd segment (though all-wood sitārs are sometimes found) and is based on the ‘large sitār’ of the 19th century, when it was standardized to a length of about 120 cm, excluding the string holder. Other dimensions are more variable. For all but very cheap instruments two types of wood are mostly used. Toonwood (tun, tūn, tunnā) has always been thought best for sound. Teak (sāgvān, sāgūn, segūn) was formerly employed only for cheaper instruments, but with the modern higher pitch and use of thicker strings (especially in Calcutta sitārs) it has become increasingly used for its strength, and combinations with a teak neck and toon soundtable are found. The body has two principal parts: the resonator or shell (khol), and the neck (dad, dandā, dandī: ‘stick’), each being composite (fig.1).
The shell has three main sections: a bulging segment (A) of about two-thirds of a gourd (tumbā, tubā, tōbā), carefully chosen for its near symmetry and acoustic properties, cut and cleaned, dried and varnished hard; a piece of thick wood (B), shaped like the half-section of the neck and shoulder of a large round bottle (gal, galā, Urdu gul, gulū, gardan: ‘neck, throat’) with an under-shelf around its lower rim on which the gourd is glued; and the soundtable (C; tablī), much thicker than in Western lutes, appearing convex but continuing the notional height of the fingerboard in its flat upper-central surface, and carved gently sloping down and away to its outer edge. On many sitārs seven carved leaves with seven points (D) project over the gourd from the rim’s lower edge.
The neck proper (E) is a long, hollowed piece of wood (its top closing wall, pagrī, ‘turban’, is usually integral), rounded at the back and roughly 90 cm long and 9 cm wide; it terminates at the lower end in a heavy tenon to which the soundtable and shoulder are nailed inside the shell. The fingerboard (F; patrī: ‘plank’) is in three pieces: over the peg area; between the string-guider (G) and the nut (H), this piece often bearing the maker’s label; and under the frets, this section being concave with narrow flat ledges running down each side.
The peg area, apart from the front, is not a separate piece on the sitār, but simply the top of the neck above the string-guider and nut (G and H); its ‘sawn-off’ shape and arrangement of frontal and right lateral pegs are distinctive features shared by the western Central Asian long-necked lutes with wooden table. The five principal pegs (khutī, kīl, kīlak: ‘peg’; Bengali also kān: ‘ear’) and their strings (tār, derived from Persian: ‘metal string’) are arranged as shown (I (i)–(v)); the strings, secured through narrow holes in the stems of the pegs, are wound anti-clockwise except for I (ii) which is wound clockwise for spacing from I (i). Modern pegs are thick and round in cross-section, with a bulbous top carved smooth, or in whorls, roses etc. for a better grip. Older sitārs have smaller pegs with a two-dimensional ‘two-leaf clover’ top. The strings are all of metal: the first and fifth are always tempered steel, the second of copper or phosphor bronze, and the others of brass or steel, according to tuning.
From the pegs the strings pass over two blades of bone or similar material set in the fingerboard and often termed ‘cross-pieces’ (ār, ārī), though their functions are different. The upper may be called the string-guider (G; Urdu tārdān, Hindi tārgahan: ‘string holder’), for it guides the strings, threaded through holes halfway down, to the nut (H; atī, from at: ‘check, restraint’), where they are held firmly in little grooves.
From the nut the main strings pass down over the frets to the main bridge (J; ghorā, ghurī, ghurac: ‘horse, mare, little horse’, respectively). This deep bridge, 7 to 7·5 cm wide, about 3 cm deep and 2 to 3 cm high, consists of antler or bone plate glued on a table-shaped wooden trestle whose two broad legs are set near the centre of the table. The surface of the plate is filed in a parabolic contour (javārī), with the node between a third and halfway back. When plucked, the string beats on the bridge in front of the node, producing a bright tone rich in harmonics; a sharper angle of filing (khulī: ‘open’), as with most sitārs, gives a twangy, nasal tone and a looser string, while a more gradual one (gol: ‘round’) gives a more subtle, veiled tone and a tighter string (as in the style of Ustad Vilayat Khan, for example). The javārī, which regularly needs renewal, serves three main purposes: it provides a long-lasting tone; it can be adjusted to allow an even timbre along the whole length of the neck (two octaves); and it reduces the tension on the string where it sits in a groove in the back wall of the bridge behind a lateral ‘ditch’ (the string would otherwise break when pulled sideways for mīnd, which can be obtained over an interval up to a 5th). The contour is often different under each string.
The strings pass from the bridge over a protective plate on the table to the inferior string holder (K), sometimes called tārdān etc. (see G above), but more often named after its shape: a long narrow triangular bone piece with an upper T-bar (langot: ‘loin cloth’; or mogrā: ‘mallet’) screwed to the gourd. Fastening devices vary from one or more projecting bone hooks around which the strings are looped in a noose, to a combination of a hook for the sympathetic strings and six or seven projecting teeth on the T-bar, one for each main string. Before being attached the melody strings are threaded through fine-tuning beads (L; mankā) which lie either on the peg area or, and in the case of the first string always, on the soundtable below the bridge, the latter usually in the form of an animal or bird; these allow rapid minor retuning during performance.
The frets (M; pardā, from Perso-Turkish pardah) are of thick curved brass, of round or oval cross-section, the two ends resting on the narrow ledges of the fingerboard and tied with bindings (fig.1b, N; bandhanī) of gut, nylon or a rough yarn of wild silk. Modern frets are side-tied, the bindings sitting in small grooves near either end of the fret and passing around the back of the neck. The 20 frets give a range of two octaves on the first string. Apart from the tonic (sa), fifth (pa), held to be immovable (acal), and the lower octave, which is chromatically complete, the second and sixth frets of the middle and the second to fourth frets of the upper octave require moving to access all the chromatic (vikrt) notes. Completely chromatic (acal-thāt) sitārs are very rare.
Also from the mid-19th century one or two thin strings began to be added, known today as cikārī (Hindi: ‘squeaking, gnat’), of 0-gauge steel and tuned variously, but often to the middle tonic (O (i)) and upper tonic (O (ii)). Deriving from the bīn through the sūrbahār, they are best described as punctuating strings, not drones. They pass from their pegs below the nut over two small bone posts (P) set upright in the fingerboard ledge (sometimes carved in the shape of cloves, and so called: laung, lavang) and down over the main bridge to the string holder. The posts are set approximately a third and two-thirds along the fingerboard; on older sitārs each cikārī peg is immediately above its post.
Another later 19th-century development on the sitār is the dozen or so sympathetic strings (taraf, tarab: ‘side-strings’), also taken from the sūrbahār, and ultimately perhaps from the sārangī. Their small pegs (Q) project sideways through the neck, and the strings rise up through a line of bone-ringed holes in the concave fingerboard, passing down under the frets to their own small bone bridge (R), also parabolically filed, and under the main bridge to the string holder. Formerly of brass, they are now made of thin steel (0 or 00) and are tuned for sympathetic resonance. Decorative features on the sitār mostly hide the joints of the parts and include a long punched strip (Hindi gōt: ‘hem’) running around the joints of the resonator and shell, recalling the inlay of Central Asian lutes; a large floral panel covering the front neck-table joint; two inlaid birds on the upper quadrants of the table; and raised wooden vines near the edge of the table in these quadrants. Sometimes the shoulder is heavily carved.
Many sitār have a small second gourd resonator (S; tumbā) attached at the back of the neck below the nut; the gourd, up to 22 cm in diameter, is capped by a round wooden shoulder, ending in a heavy tubular brass screw inserted in a nut built into the neck. This is detachable, and is not an original feature of the sitār; it adds little extra resonance and serves perhaps only a symbolic purpose, manifesting a bīn lineage for the player.
The sitār is always played with a twisted-wire plectrum (Urdu mizrāb), worn on the right index finger (fig.1c).
This is effectively a bass sitār. It was invented about 1820 by the famous sitār player Ghulam Muhammad of Lucknow (or, some say, by his teacher, the bīn player Pyar Khan) as an instrument suitable for playing the older Hindustani style of the bīn. Its construction is essentially that of the sitār, but with the following differences: the overall dimensions are much larger, with a length of 145 cm or more, a neck width of at least 11 cm and the diameter of the soundtable over 40 cm; the gourd-section at the back of the shell is flat-backed and round-sectioned (as is the table), often with a projecting wooden floor-rest on the left side of the gourd; the tied curved metal frets are often vertically flat-sectioned, with small flat plates at either end for support; the pegbox is separate, bent-back and often has a scroll, open at the back and with a bilateral (two left, three right) arrangement of the main pegs. Some sūrbahār have also a soundhole on the table, or a second gourd resonator.
It is still mainly performed by sitār players, with the same plectrum and technique (though some use that of the bīn). Its Urdu name means literally ‘a springtime of notes’, referring to the sympathetic strings, then unusual on long lutes. It is considered to have three ‘breaths’ (dam), in that the dying sound can be revived twice by left-hand portamento (mīnd), which in the hands of masters (of which there are but few) can extend to a full octave.
The player sits fully on the floor, his left leg tucked flat beneath his right, the shell supported in the hollow of his left foot. The main weight of the sitār is in the neck, and the shell must be held down by the right forearm, hanging naturally. The left thumb maintains a steady pressure on the side of the neck, with an angle at the wrist; the knuckles press down over the first string. Some players sit cross-legged with the raised right thigh supporting the neck (requiring less pressure from the right arm); others keep it flat.
The basic plucking style is a continuous through-movement of the right hand from the wrist, all four fingers held loosely together and supported by the thumb, mostly on the first string but often lightly brushing the second string simultaneously. Only the plectrum is used for plucking, with the hand tilted somewhat, generally inwards (dā) for a downbeat and outwards (rā) for an upbeat. The left hand cups the neck lightly, touching it only at the back with the rigid thumb and on the frets primarily with the gently curved index finger between the pad and the tip; this light grip is maintained throughout performance, the middle finger touching for a descending turn, or the third for a large stretch (but some players ascend with the middle finger and descend with the index, or vary their fingering according to context). The constant use of one finger creates the Indian vocal legato quality, and the distance between frets is too great for much use of positional fingerings.
Full notes on the frets are ‘standing notes’ (Hindi kharā sur), but there are many ornamental techniques, of three main types. The first kind are played along the string: in ghasīt (‘dragging’) the finger slides to another fret, lightly for grace-notes (kan: ‘drops’), up or down (Bengali biksep, praksep: ‘casting’) or slow (gharsan: ‘friction’ or ās: ‘expectation’) to suggest intermediate tonal-microtonal pitches; a dramatic slide is chūt (‘release’). Hammering with the middle finger is sparś (‘touch’), or kan, and repeated hammering zamzamā (Urdu: ‘humming’); in krntan (‘cutting’) the middle finger plucks off a fret. The second type consists of a sideways pulling (khic) of the string for portamento (mir, mīnd) up to a 5th or, rarely, a flat 6th, derived from the vīnā but, like the sūrbahār, pulling to the left of the neck, free of the strings. The sitār is considered to have two dam in mīnd. Āndolan is a slow repeated ‘swing’, and garnak a throbbing fast one. The third kind, a cross-string plucking (cher, cher-chār), is principally a feature of jhālā since the adoption of cikārī strings.
K.M. Gosvami: Sangīt-sār (Calcutta, 1868) [in Bengali]
S.M. Tagore: Yantra-ksetra-drpīkā (Calcutta, 1872) [in Bengali]
S.M. Tagore: Yantra-koś (Calcutta, 1875/R) [in Bengali]
K.D. Banerji: Gīta-sūtra-sār (Calcutta, 1885) [in Bengali]
B.K. Roy Choudhury: Bhūratiya-sangīt-koś (Calcutta, 1965) [in Bengali]; (New Delhi, 1975) [in Hindi]
B.S. Sarma: Sitār-mālikā (Hathras, 3/1966) [in Hindi]
J.N. Pathak: Sitār-siddhānt (Allahabad, 1967) [in Hindi]
R. Shankar: My Music, My Life (New York, 1968/R) [incl. manual on playing the sitar]
A. Khan: Sitār-darpan (Baroda, 3/1972) [in Hindi]
M. Junius: The Sitar (Wilhelmshaven, 1975)
L. Sharma: ‘Development of the Sitar in India’, Journal of the Music Academy, Madras, lvii (1986), 205–12
S. Slawek: Sitar Technique in Nibaddh Forms (Delhi, 1987)
S. Bandyopadhyaya: Techniques of Sitar (Delhi, 1988)
J.S. Hamilton: Sitar Music in Calcutta (Calgary, 1989)
A. Miner: Sitar and Sarod in the 18th and 19th Centuries (Wilhelmshaven, 1993)
