Iacobus Leodiensis [Iacobus de Montibus, Iacobus de Oudenaerde]

(ii) Improvisation on compositions.

Not only the underlying form of vocal compositions but also the basic types of nibaddha (‘constrained’) improvisation are prefigured in the Sangīta-ratnākara. The elaboration of a metrical composition (rūpakālapti) is said by Śārngadeva to comprise either of two processes: pratigrahanikā (‘reprising’) rūpakālapti, where after a phrase of rāga elaboration one returns to a phrase of the composition; or bhañjanī (‘breaking up’) rūpakālapti, where the text of the composition is sung many times, in part or whole, keeping the original rhythmic setting of the words but with ever new melodic phrases. Similar processes are still current in modern practice; in particular, the alternation of improvised episodes with periodic reprises of a phrase of the composition, usually the first, is a ubiquitous practice.

Śārngadeva's discussion of rāgālapti (see §3(ii)(b) above) and rūpakālapti in the Sangīta-ratnākara provides us with the first analysis of the processes of improvisation in South Asian music. There is also an indication of how these processes related to performance as a whole, in the description of a courtly dance-form, the Gaundalī (‘Gond woman's dance’). According to this description the performance of this dance began with the singing of the tonic (sthāyī) and the fourfold development of rāgālapti. A vocal composition (dhruvā-prabandha, see above) was then sung and danced in tāla with drum accompaniment. New melodic material (sthāya) was then introduced, interspersed with reprises of the refrain of the composition (dhruvā-khanda). The reprise was marked by vigorous drum playing and a sudden ‘freeze’ by the dancer at the crucial moment of return, a process familiar from modern kathak dance performances. This sequence of events – rāgālapti, composition, and (pratigrahanikā) rūpakālapti punctuated by returns to the refrain – underlies the major genres of rāga performance in both Karnatak and Hindustani music today.

(a) Returns to a given motif (pratigrahanikā).

Here a particular phrase of the composition, often the first, is repeated in alternation with passages of improvised melodic and/or rhythmic development. This process is used in all the styles in the ‘nibaddha’ column of Table 9 above. It underlies the bol-bat of dhrupad, the ālāp and tāns of khayāl, the bol banāo of thumrī, the torā of sitār /sarod playing and the kalpana svara of Karnatak music. Whatever the style and techniques employed, musical interest lies both in the melodic and rhythmic excursions of the improvisation and in the ingenuity with which the performer returns to the composition. The Hindi term for the motif to which the return is made is mukhrā (‘little face’); it is normally the first phrase of the sthāyī. In ex.3 above, for instance, ‘sapane-mẽ ā-’ is the mukhrā; it has two parts, comprising the point of arrival at ‘ā’ and the phrase leading up to that point, which is, properly speaking, the mukhrā. The point of arrival is called sam (‘together’), where both melodic and rhythmic elements finally coalesce at a single point. In ex.2 above the mukhrā is ‘hajarata’, and the sam is on ‘to’. As is illustrated here, in Hindustani vocal music, and especially in slow khayāl, a precisely measured-out duration is not always essential to the mukhrā, the musical function of which is as a lead-in or lead-back. In rendering the slow khayāl in rāga Darbārī Kānadā, for instance, phrases of unpulsed improvised ālāp would fill most of each long time cycle (transcribed here as 12 whole-note values); the singer will slip into a pulsed phrase as he comes back to ‘hajarata’, making sure only to catch it far enough ahead of the sam to allow time for an elegantly shaped approach to and arrival at ‘to’.

The tān of Hindustani fast khayāl are free bravura passages sung on an open vowel. In tān singing, some flexibility as to the amount of the mukhrā actually used or the point at which it begins, or both, is again possible; the tān must conclude and some rhythmically elegant mukhrā should be heard, but the only irreducible metric requirement is that the sam be reached with the first count of the time cycle. Sometimes tān may lead right to the sam, dispensing with the mukhrā altogether. The mukhra can be treated with similar flexibility in dhrupad (and especially dhamār) laykārī improvisation (see below).

In Hindustani instrumental music the principal type of free passage between returns of the mukhrā is called torā (‘break’, also ‘bracelet, necklace’). A torā is a plucking pattern for the right hand combined with pitch-changing movements of the left hand. A torā differs from a tān in that it has some sort of definite rhythmic shape, whereas a tān is passage-work in fast but equal note values. A torā often concludes with a culminating cadential passage repeated three times in sequence, called tihāī (‘threefold’). The tihāī must be timed in such a way that it concludes on the beat before the mukhrā, or at the beginning of the mukhrā, or the sam itself. Tihāī can also be used in dhrupad and is an important feature of dance performance.

The flexibility of beginning point and emphasis on arrival point in the Hindustani mukhrā is reversed in Karnatak music. A line of the composition (not necessarily the first) is chosen, and free improvisation is required to lead directly, naturally, but also precisely to the first part of the phrase, which then may or may not proceed to a strong arrival point (arudi). Ex.14 illustrates one of the characteristic specialities of Karnatak music. It shows sample phrases of kalpana svara (‘improvised svara’) attached to the first line of the composition partly transcribed in ex.9, Vēnugānalōluni, rāga Kedāragaula (see §2(iv) above). The firmly fixed point is the actual beginning of the line of the composition, ‘Vēnu-gāna-’, which is called its etuppu (‘taking up’). The etuppu is a specific svara of the rāga, falling at a specific point in the tāla; here it is the second degree (d) in rāga Kedāragaula, coming on the second half of the first beat in fast-tempo of Rūpaka tāla. There are only a few ways this point can be approached in Kedāragaula; in fact, the last three degrees of a kalpana svara passage could lead to this motif beginning at d only by going e–d–c–e (as in ex.14), g–f–e–d or f–e–c–d (following g or d). The time-point for the etuppu here, just after the first count, is the norm for fast Rūpaka tāla; the final approach to the etuppu with an odd number of attacks (five in this and most cases) is the norm for all tālas. The low initial pitch and immediate upward continuation of the composition's opening demand a basically descending line in each final approach. The requirements of the rāga Kedāragaula (see §2(iv) above) determine the specific possibilities of melodic configuration in the kalpana svara throughout.

Two general techniques of pratigrahanikā improvisation are illustrated in ex.14. First of all, a fixed formula at the end is preceded by an ever longer and more elaborate build-up. Also characteristic is the use of the threefold cadential rhythmic sequence at the ends of longer passages, called morā in Karnatak music. In the transcription's semiquaver durations, the third of the sample kalpana svara ends sequentially with unit durations of 4 + 4 + 5 pulses. The fourth passage ends 6 + 6 + 5, with a rhythm that is one of many possible rhythms for a standard mrdangam cadential formula (tadinginatom, see §6 below). The drummer would probably be expecting this formula and would certainly join in in time for the second and third members of the sequence. (Hindustani musicians sometimes improvise with svara syllables and with other non-textual syllables (bol) in compositions that use them, but seldom with tight linkage to a precise time-point or the control of rhythmic pattern that is essential to the modern south Indian technique.)