PART III

DRAMATIC THEORY

XIII

THE THEORY OF THE DRAMATIC ART

1. THE TREATISES ON DRAMATIC ART

Pāṇini, whose date falls doubtless before 300 B.C., alludes in his grammar to the Naṭasūtras, books of rules for Naṭas, compiled by Çilālin and Kṛçāçva, and Professor Hillebrandt [597] has suggested that we should recognize in these works the earliest text-books of the Indian drama. But we have no other suggestion that Pāṇini knew of dramatic performances, and the only legitimate conclusion is that these rules were laid down for the guidance of dancers or, perhaps, pantomimes, and with this accords admirably the fact that the dramatic tradition knows nothing of these names, and instead makes the sage Bharata the eponymous hero of the drama. True it was Brahmā, highest of gods, himself who, at the instance of the gods, produced as a counterpart to the four Vedas, which contain the science of religion and magic, the more mundane Nāṭya-Veda, consecrated to the drama, but this Veda is not current among men. Bharata, on the other hand, whose task it was to direct the production by the Apsarases in heaven of plays for the delight of the gods and who thus had practical experience of the art, has set forth for men the principles of the drama in the Nāṭyaçāstra which, if not inspired, has at least a measure of sanctity, and thus supplies an authoritative basis for practice.

The legend is interesting because it precisely interprets the spirit of India towards authority; Bharata occupies in the theory of the drama a place analogous to that of Pāṇini in grammar, but unfortunately the Nāṭyaçāstra has fared badly in comparison with the Aṣṭādhyāyī, which has, through the care of its commentators, come down to us in a form but little changed from that it assumed in the hands of its author. The work, which we have under the title Bhāratīya Nāṭyaçāstra, [598] is extremely badly preserved in the manuscript tradition, a fact due in part to the comparatively late date of any commentary upon it. We have only a few references to an exposition of the Nāṭyaçāstra by Mātṛgupta, a somewhat mysterious figure with a more or less legendary connexion with Kālidāsa, with whom he has even been identified; [599] if we are to place any faith in his contemporaneity with Kālidāsa, he may date from the close of the fourth century A.D. It is significant that tradition makes him for a time king of Kashmir, for it is to that country we owe the commentaries of Çan̄kuka, who wrote the epic Bhuvanābhyudaya under Ajitāpīḍa (A.D. 813–50), and of Bhaṭṭa Nāyaka, who belongs to the period of Çan̄karavarman (A.D. 883–902). In the same line of tradition is the great work of Abhinavagupta, the Abhinavabhāratī, which has been lucky enough to come to light after long oblivion, and which represents the erudition of the close of the tenth century.

The treatise, as we have it, is elaborate, covering the whole ground connected with the drama. It deals with the architecture of the theatre, the scenery, and the dress and equipment of the actors; the religious ceremonial to be observed at every representation; the music, the dance, the movements and gestures of the actors, and their mode of delivery; the division of rôles; the general characteristics of poetry; the different classes of drama, and the emotions and sentiments which form a vital element in the drama. There is confusion, complexity, and repetition in the work, but that much of it is old cannot be doubted. It appears clearly to be based on the examination of a dramatic literature which has been lost, eclipsed by the more perfect dramas of Kālidāsa and his successors. In the description of classes of drama we seem to have hasty generalizations on insufficient material; the Samavakāra, for instance, is described in terms which, with the precise definition of the time to be occupied by the acts, can be interpreted only as based on a single drama, and the Ḍima seems to have a similar origin. The elaborate description of the preliminary scene or Pūrvaran̄ga, which is practically non-existent in the classical drama, suggests a period of a less cultivated taste. A more definite result may be derived from comparison of the Nāṭyaçāstra with the works of Açvaghoṣa and of Bhāsa. The Prākrits recognized by the Nāṭyaçāstra are clearly later than those of Açvaghoṣa and more akin to those found in Bhāsa; again the Nāṭyaçāstra recognizes the use of Ardha-Māgadhī, found in these two dramatists, but not later, while, like them, he ignores the Māhārāṣṭrī of the later dramas. Moreover Bhāsa expressly alludes to a Nāṭyaçāstra [600] and it is most probable that both he and Kālidāsa had knowledge of the prototype of the present text. That Bhāsa by no means slavishly adheres to the rules of the Nāṭyaçāstra, either as regards the formal mode of terminating his dramas or the exclusion of scenes of death from the stage, [601] merely shows that when he wrote the Çāstra had not attained any binding force. There is nothing to contradict the date thus vaguely indicated, [602] for the treatment of poetics in general is simple and early, and it is impossible to draw any conclusion as to date from the remarks on music, apart altogether from the constant possibility that incidental additions and alterations have been made in the work.

It was inevitable that the complicated and confused work of Bharata should be superseded for many purposes by something more accessible and easy to follow, and this need was supplied by the Daçarūpa of Dhanaṁjaya, son of Viṣṇu, and protégé of the ill-fated king Muñja of Dhārā (974–95). The work takes its name from the ten primary forms of drama recognized in the Nāṭyaçāstra, which is followed closely by Dhanaṁjaya, his deviations being unimportant and trivial, such as a new division of types of heroine or of the erotic sentiment. On the other hand, Dhanaṁjaya omits by far the greater part of the topics of his model; his four books of wooden verses treat first of the subject-matter and plot; then of the hero, the heroine, and other characters, and the language of the drama; thirdly of the prologue and the different kinds of drama; and lastly of the emotions and sentiments, thus concentrating attention on the essential dramatic features. The text is naturally often unintelligible save in the light of the Nāṭyaçāstra itself and of the commentary, Avaloka, which is ascribed to Dhanika, son of Viṣṇu, and minister of Utpaladeva, a term which is an alias of Muñja. The identity of the two writers is suggested by the fact that later writers ascribe passages of the Daçarūpa itself to Dhanika, and that without the commentary the work is in a sense incomplete. But, on the other hand, in a few passages the commentator more or less distinctly differs from the text, and it seems sufficient to assume that they may have been brothers. The Avaloka must have been completed after Muñja’s death, since it cites Padmagupta’s Navasāhasān̄kacarita, which was written under Sindhurāja, and this throws some doubt on the identification of Dhanika with the Dhanika Paṇḍita to whose son, Vasantācārya, a land grant was made by Muñja in A.D. 974. Dhanika quotes stanzas of his own in Sanskrit and Prākrit and also a treatise, Kāvyanirṇaya, elsewhere unknown. [603]

Of the fourteenth century in all probability are three works of unequal importance and merit. The Pratāparudrīya [604] of Vidyānātha is a mediocre compilation from the Daçarūpa and the Kāvyaprakāça of Mammaṭa, covering the whole field of poetics; it illustrates the formal rules of the drama by the composition of a wretched drama in honour of Pratāparudra of Warangal, whose inscriptions show dates from A.D. 1298 to 1314. Of much greater interest is Vidyādhara’s Ekāvalī; [605] like Vidyānātha, the author celebrates in his illustrations of his text his patron, in this case Narasiṅha II of Orissa, perhaps A.D. 1280–1314; as a poet his merits are negligible, but he shows a lively interest in his subject and intelligence in his views. Of greater popularity than either is Viçvanātha Kavirāja, the author of the Sāhityadarpaṇa, [606] a general treatise on poetics. His handling of the drama is based largely on the Daçarūpa and its commentary, but he introduces a good deal of matter from the Nāṭyaçāstra in his sixth chapter, including details of the characteristics and ornaments of the drama, which the Daçarūpa omits. In this Viçvanātha indicates his servile character, which, however, renders his work the more valuable as an exposition of the orthodox doctrine. Of his ancestry and his own works he makes free mention, but the most definite evidence of his date is the existence in the library at Jammu of a manuscript of his work whose date appears to be A.D. 1383. The lack of order and the errors in his work are made the basis of criticism by Rūpa Gosvāmin in the early part of the sixteenth century, but his own Nāṭakacandrikā shows little improvement on the work of his predecessor, whence it draws much of its material; its real purpose is to eulogize the saint Caitanya, whose disciple Rūpa was and in whose honour he composed dramas of no merit. Equally dependent on Viçvanātha and the Daçarūpa is Sundaramiçra, whose Nāṭyapradīpa was composed in A.D. 1613. Many other treatises on drama are known by name or exist in manuscript, but none apparently of any great importance or repute. Of the fourteenth century also is the Rasārṇavasudhākara [607] of Çin̄ga Bhūpāla, lord of Rājācala and the land between the Vindhya and Çrīçaila about A.D. 1330, who cites Vidyādhara.

The development of a theory of drama progressed in the closest relation to the general theory of poetics, for the Indian theory of poetry does not admit any distinction in essence between the aesthetic pleasure produced by the drama and any other form of poetry. Thus we find in Abhinavagupta in full application to the drama the theory of suggestion, Dhvani, as the essence of poetry, which appeared in strength about A.D. 800 and was rendered popular by Ānandavardhana (A.D. 850) and by Abhinavagupta himself in his comment on the Dhvanyāloka of the former. Attacked by Mahiman Bhaṭṭa, author of the Vyaktiviveka (A.D. 1050), the doctrine was again developed with special care by the Kashmirian Mammaṭa [608] at the close of the eleventh century. In slightly varied forms it appears in Vidyānātha, Vidyādhara, and Viçvanātha.

Apart from this important development, which, however, has no special application to the drama, there is little progress in the course of the literature. The later authorities are bound by the authority of the Nāṭyaçāstra; they repeat unintelligently its descriptions of literary forms such as the Ḍima, the Samavakāra, the Īhāmṛga, the Vīthī, and the An̄ka, which had ceased to be in popular use, if indeed the definitions of the Nāṭyaçāstra were not merely hasty generalizations from a single play or so in every one of these cases. The most that they do is to omit or to vary details, but not in independence; normally the changes can be traced to variants in the text of the Çāstra or to maxims current under Bharata’s name, though not included in the Çāstra as we have it. Often the authors differ in the definition of terms in the Çāstra which, as often in Sanskrit technical phrases, present ambiguity and admit of various renderings. These divergences are especially frequent in the long lists of characteristics and ornaments or the different means of effecting dramatic results; the Indian love of meaningless subdivision here can indulge itself to its fullest and least profitable extent. A rich variety of such ambiguities is apparent in the verses in which the Agni Purāṇa [609] describes the drama, including dancing and the mimetic art, true to its aim to constitute itself a treasure-house of all learning, popular as well as divine. The chief value of the work is the occasional light which it throws on the variants in the text of the Çāstra, and its comparative antiquity, for it is cited in the Sāhityadarpaṇa and is probably some centuries older.

2. THE NATURE AND THE TYPES OF THE DRAMA

A drama is the imitation or representation of the conditions or situations (avasthānukṛti) [610] in which the personages who form the subject of treatment are placed from time to time, by means of gesture, speech, costume, and expression, and, one version of the definition adds, the situations must be such as to produce pleasure or pain, that is, they must be tinged with emotion. It is the presence of these ancillaries which distinguishes the drama from an ordinary poem; a poem appeals to the ear only, a drama is also a spectacle to delight the eyes; hence the term Rūpa or Rūpaka as applied generically to the drama, for Rūpa primarily denotes the object of vision, though the Indian tradition gives the artificial explanation that Rūpaka denotes a drama because the actors are credited with different parts.

Further light is shown on the nature of drama (nāṭya) by the discrimination of it from dance (nṛtta) and mimetic art (nṛtya), which united with song and speech serve to make up the drama. [611] The dance is based on time and rhythm; the mimetic art is concerned with representing the feelings or emotions (bhāva), while the essence of the drama is the sentiment (rasa) which it evokes in the spectator, a fact which places it on a higher level than either of its handmaidens. But there may be dramas in which these auxiliaries take first place, and on this fact is based a distinction between the primary forms, Rūpakas, in which the poetry is the dominant element and the secondary forms, Uparūpakas. Of Rūpakas ten are distinguished, Nāṭaka, Prakaraṇa, Bhāṇa, Prahasana, Ḍima, Vyāyoga, Samavakāra, Vīthī, An̄ka, and Īhāmṛga, which vary in regard to subject-matter (vastu), hero or heroine, and sentiment.

3. THE SUBJECT-MATTER AND THE PLOT

The scene of the plot must be laid in India, and the period must be one of the three ages succeeding the Golden Age, for pleasure and pain, essential elements as we have seen in the drama, cannot be experienced elsewhere than in Bhāratavarṣa, and even there they do not exist in the age of happiness unalloyed. [612] Otherwise the choice is free; the poet may take an incident familiar from tradition (prakhyāta), or may invent his plot (utpādya) or may combine both forms (miçra). But, if he follows a current legend, it is necessary that he shall not ruin the effect of it by incongruous invention; he must confine his ingenuity to episodes, for otherwise the audience will be painfully disturbed by departure from tradition. On the other hand, it is not merely legitimate but also necessary that the dramatist should ennoble his hero, if tradition assigns to him deeds incompatible with the character which he normally exhibits. [613] The epic is not encumbered with such considerations; it can represent Duḥṣanta as merely forgetful of his vows to Çakuntalā, but Kālidāsa must clear the character of the king from this seeming baseness by attributing his loss of memory to a curse provoked by a negligence of the heroine herself. The Rāmāyaṇa admits, and seeks to explain, if not convincingly, the death of Vālin, king of the monkeys, at the hands of the virtuous Rāma; Māyurāja in the Udāttarāghava passes over the episode in silence, while Bhavabhūti, with greater boldness, in the Mahāvīracarita perverts tradition to represent Vālin as an ally of Rāvaṇa, and as slain by Rāma in legitimate self-defence, and exonerates Kaikeyī.

The subject-matter takes two forms, the principal (ādhikārika) and the incidental (prāsan̄gika) actions. The first owes its name to the fact that it is connected with the attainment (adhikāra) of the purpose of the hero, whether that be love, or some material interest, or duty, or two or all of these. In the incidental action the end achieved is not that aimed at by the hero, but it serves as a means towards the fruition of his aims. [614] The incidental action may take the dimension of an episode (patākā), as is the case with the exploits of Sugrīva as an ally of Rāma, or it may be a mere incident (prakarī), as in Act VI of the Çakuntalā the scene in which the two attendants converse. [615]

An action, when developed in full, as normally it is in the Nāṭaka, the most perfect of forms of drama, involves of necessity five stages of development (avasthā); [616] there must be as the beginning (ārambha) the desire to attain some end, which leads on to the determined effort (prayatna) to secure the object of desire; this leads to the stage in which success is felt to be possible (prāptyāçā, prāptisambhava) having regard to the means available and the obstacles in the way of achievement; then arrives the certainty of success (niyatāpti), if only some specific difficulty can be surmounted; and finally the object is attained (phalāgama). Thus in the Çakuntalā we have the king’s first anticipation of seeing the heroine, then his eagerness to find a device to meet her again; in Act IV we learn that the anger of the sage, Durvāsas, has in some measure been appeased, and the possibility of the reunion of the king and Çakuntalā now exists; in Act VI the discovery of the ring brings back to the king remembrance, and the way for a reunion is paved, to be attained in the following act. The Ratnāvalī, no less perfect an example of the minor type, the Nāṭikā, reveals to us the aims of the minister to secure the union of the heroine and the king; a definite step to this end is taken when the heroine decides to depict the face of Vatsa on the canvas; in Act II the lovers are united for the moment, but subject to the risk of discovery by the queen; then the king recognizes that his success in love depends on winning the queen’s favour, which is successfully accomplished in the last act.

There are also five elements of the plot (arthaprakṛti), [617] which the theory not very accurately parallels with the five stages of the action. The first is the germ (bīja) whence springs the action, as in the Ratnāvalī from Yaugandharāyaṇa’s scheme to secure the princess for the king. The second, with change of metaphor, is the drop (bindu) which spreads out as oil on water; the course of the drama, which has seemed to be interrupted, is again set in activity; thus in the Ratnāvalī, when the festival of the god of love is over, the princess gives a decisive impulse to the motion of the drama by recognizing in him, whom she deemed the god himself, the king for whom she was destined as a bride. The other three elements are the episode, the incident, and the dénouement (kārya).

Based on these parallel sets is a third division of the junctures [618] (sandhi), which carry each of the stages of the action to its natural close. They are the opening (mukha), progression (pratimukha), development (garbha), pause (vimarça), and conclusion (nirvahaṇa), corresponding clearly and closely with the stages [619] set out above. Thus in the Çakuntalā the opening extends from Act I to the point in Act II where the general departs; the progression begins with the king’s confession to the Vidūṣaka of his deep love, and extends to the close of Act III. The development occupies Acts IV and V, up to the point where Gautamī uncovers the face of Çakuntalā; at this moment the curse darkens the mind of the king, who, instead of rejoicing in reunion with his wife, pauses in reflection, and this pause in the action extends to the close of Act VI, while the conclusion is achieved in the last Act. In the Ratnāvalī the opening extends to that point in Act II, where Ratnāvalī decides to depict the king as the only means of gazing on him whom she loves, but from whom she is jealously kept by the queen; the progression extends then to the close of the Act; the development occupies Act III, while the pause, due to the intervention of the queen, is brought to an end by the mock fire of the palace in Act IV, and the remaining portion of that Act gives the conclusion.

So far there is obviously force and reason in the analysis, which, if in needless elaboration, recognizes the essential need of a dramatic conflict, of obstacles to be overcome by the hero and heroine in their efforts to secure abiding union. The classification of elements of the plot is perhaps superfluous beside the junctures; its parallelism to the other two divisions is faulty, for it is admitted that the episode is not confined to the development, as it should be, but may extend into the pause and even into the conclusion. [620] The episode again is credited with sub-junctures, to be fewer in number than the junctures, and even the incident is permitted on one view to have incomplete junctures. [621] But far more complex is the insistence on the subdivision of the five junctures into 64 members (12, 13, 12, 13, and 14 respectively). The distribution, however, has no real value, for, though Rudraṭa [622] asserts that the members should only be used in the juncture to which they are assigned, other authorities decline to admit this view, on the score of the usage of the dramatists, which is the supreme norm. Not all of these members need be used; it is a fault in the Veṇīsaṁhāra that the poet drags in the separation of Duryodhana from Bhānumatī in Act II for no better reason than to comply with the rules. [623] When used, they should be essentially subservient to the sentiment which the piece seeks to create; [624] they should either treat the subject chosen, expand the plot, increase interest, produce surprise, represent the parties in action, or conceal what should be concealed; the hero or his rival should appear in them, or at any rate they should flow from the germ and lead up to the dénouement. Some must be included in any drama, since one without any would be like a man without limbs, and, adroitly used, they may give merit to a mediocre subject-matter. But the definitions and the classifications are without substantial interest or value.

A distinction must be made between such things as can properly be shown on the stage, and such as must only be alluded to. [625] What is seen should essentially serve to produce the sentiment aimed at, and it must avoid offending the feelings of the audience. Hence it is improper to portray on the stage such events as a national calamity, the downfall of a king, the siege of a town, a battle, killing, or death, all of them painful. It is equally forbidden to depict a marriage or other [626] religious rite, or such domestic details as eating, sleeping, bathing, or anointing the body, amorous dalliance, scratching with nails or teeth, or such ill-omened things as curses. But these rules are not without exception early or late; if Bhāsa does not hesitate as in the Ūrubhan̄ga to depict death on the stage, Rājaçekhara in his Viddhaçālabhañjikā describes the marriage ceremonial in Act III, and in the following Act shows us the wife of Cārāyaṇa asleep, while the author of the Pārvatīpariṇaya does not hesitate to choose as his theme the nuptials of Çiva and Pārvatī. Nor do dramatists decline to represent death if the dead person is restored to life, as in the Nāgānanda. [627] A long journey, or calling from a distance, [628] is excluded from representation for obvious reasons of practicability.

Such matters as are appropriate for presentation must be presented in Acts, and each Act must contain only such events as can naturally, or by skilful management, be made to occupy the duration of a single day, [629] a requisite which is obeyed by Bhavabhūti in his Mahāvīracarita and by Rājaçekhara in his Bālarāmāyaṇa despite the difficulties presented by the effort thus to condense the epic. But it is essential that the events described shall not be disconnected; they must flow from the same cause, or issue naturally from one another. There should be an effective development of the plot within the Act; at the time when it comes to an end by the departure of the actors—three or four at most, one of whom should be the hero—at the moment when they seemed to have attained their immediate aims, a new motive should come into play, and a fresh impetus be given to the movement of the drama. But it is neither necessary nor usual that Act should follow Act without interval; on the contrary, anything up to a year may intervene between the action of one Act and that of the next; if the events as recorded in history covered more than that time, as in the case of Rāma’s fourteen years of banishment in the forest, the poet must reduce the period to a year or less. To reveal to the audience the events during such intervals the theory permits a choice of five forms of scenes of introduction (arthopakṣepaka), which serve also to narrate things, whose performance on the stage is forbidden by the etiquette of the drama. [630]

Two of these are the Viṣkambha or Viṣkambhaka and the Praveçaka, which are both explanatory scenes, but between which the theory draws fine distinctions. The Viṣkambhaka is performed by not more than two persons, [631] never of chief rank; it serves to explain the past or the future, and it may be used at the beginning of a drama where it is not desired to arouse sentiment at the outset. It is pure (çuddha) if the performers are of middle rank and speak Sanskrit; mixed (saṁkīrṇa) when the characters are of middle and inferior class and use also Prākrit. The Praveçaka cannot be used at the beginning of a drama, and is confined to inferior characters, who use Prākrit. Thus in the Çakuntalā Act III is introduced by a Viṣkambhaka, in which a young disciple of the sage Kaṇva tells us in Sanskrit of the king’s stay at the hermitage, while in Act VI a Praveçaka gives the episode of the fisherman and the police. An abbreviated mode of producing the same result is the Cūlikā, [632] in which a voice from behind the curtain narrates some essential event, as when in Act IV of the Mahāvīracarita we learn thus of the defeat of Paraçurāma by Rāma. In the An̄kamukha, or anticipatory scene, at the close of one Act a character alludes to the subject of the following Act; thus at the end of Act II in the Mahāvīracarita Sumantra announces the arrival of Vasiṣṭha, Viçvāmitra, and Paraçurāma, and these three open Act III. A different view is taken by Viçvanātha, who makes it out to be a part of an Act in which allusion is made to the subject-matter of the following Acts and the whole plot, as is done in the dialogue of Avalokitā and Kāmandakī in Act I of the Mālatīmādhava. This is evidently an attempt to justify the treatment of this form of scene as revealing matters which cannot conveniently be depicted on the stage, as well as to distinguish it from the An̄kāvatāra or continuation scene, in which the action is continued by the characters in the next Act without any break, other than the technical one of the departure of the actors and their return, as at the close of Act I of the Mālavikāgnimitra. Such a scene obviously in no way answers the purpose of explanation, and its assignment to such an end is clearly erroneous.

Various devices are recognized to help the movement of the intrigue, five of which are classed as internal junctures (antarasandhi). [633] The first of these is the dream, as in the Veṇīsaṁhāra where Bhānumatī is terrified by a vision in which she sees an ichneumon (nakula) slay a hundred snakes, dread presage of the fall of the hundred Kauravas before the attack of Nakula and his brothers. The letter serves in the Çakuntalā, Act III, to allow the heroine to express her feelings towards the king; she reads it aloud and he overhears it and breaks in upon her; more often it serves the important end of conveying news, leading to dramatic action. A message serves the same end, as when in the Çakuntalā, Act VI, Mātali brings to the king Indra’s message imploring aid against the demons. A voice from behind the scene (nepathyokti) in Act I of that play warns Duḥṣanta not to kill the gazelle of the hermitage, and a voice in the air (ākāçabhāṣita) in Act IV makes known to Kaṇva on his return the important news of Çakuntalā’s marriage and approaching motherhood. The Nāṭyaçāstra [634] ignores the term internal junctures, but has the term special junctures or divisions of junctures (sandhyantara) which includes the dream, the letter, and the message, among many other miscellaneous elements; two of these are akin to those already mentioned. The picture is used in the Ratnāvalī as the mode by which the heroine satisfies her longing for her beloved, while Vāsavadattā discovers Vatsa’s infidelity through seeing the portrait of Sāgarikā, painted beside that of the king by the mischievous Susaṁgatā. Intoxication (mada) may result as in the Mālavikāgnimitra, Act III, in the letting fall of imprudent words by an important character. Other devices might have been included in the list, such as that of assuming a disguise on the stage, a device used by Harṣa in the Ratnāvalī and the Priyadarçikā in order to secure the inconstant king uninterrupted interviews with the objects of his temporary affections. The latter play contains in Act III a good example of the embryo Act (garbhān̄ka), [635] which is recognized by the theory but not classed as a species of juncture; in it Vāsavadattā causes her maids of honour to perform before her a play representing her early adventures with Vatsa. So in the Uttararāmacarita Vālmīki has performed by the Apsarases before Rāma and Lakṣmaṇa the adventures of Sītā since her banishment, and the events of her marriage are described in this form in the Bālarāmāyaṇa, Act III.

Similarly the theory recognizes as a separate element the pro-episode (patākāsthānaka), [636] an equivocal speech or situation which foreshadows an event whether near at hand or distant. The Nāṭyaçāstra distinguishes four kinds of equivoke. An ambiguous situation may result in bringing about the aim of the hero; thus in Act III of the Ratnāvalī, when Vatsa hastens to save Vāsavadattā, as he thinks, from hanging herself, he finds to his equal joy and surprise that he has rescued none other than Sāgarikā herself. [637] Or the equivocation may lie in words, whose sense the spectator alone grasps in its deeper application; thus in Act II of the Çakuntalā a voice behind the scene bids the female Cakravāka say farewell to her spouse, a command whose application to the case of the king and the heroine is immediately appreciated by the audience alone. Or the equivocation may be deliberately conveyed in the response of the actor, whose words apply not merely to the immediate matter in hand, but allude to the future; in the Veṇīsaṁhāra, Act II, Duryodhana is told of the mishap of the breaking of his standard by the fierce (bhīma) wind in words which presage his own fall, his thigh broken by Bhīma’s blow. Finally we may have a double entendre which later is destined to find a third application; in the Ratnāvalī Vatsa playfully suggests that his earnest gaze on the creeper, which has borne blossoms out of season, may cause jealousy in the queen; his words apply equally to a maiden, and in the sequel the queen is made furiously angry by his ardent gaze at Sāgarikā. The Daçarūpa contents itself with two species, equivocation of situation and deliberate equivocation of phrase, but there is general agreement that pro-episodes may be used in any part of the play and not merely in the first four junctures.

Importance attaches to the conventions which enable the author to surmount difficulties inseparable from the dramatic form. [638] Normally, of course, the actors speak aloud (prakāçam), to be heard by all those on the stage as well as by the audience, but asides (svagatam, ātmagatam) are frequent, meant to be heard by the audience alone. If the need arises for making a remark to be heard by one actor only, it is made in the form of a confidence (apavāritam, apavārya), while a private conversation (janāntikam) is arranged by the actors holding up three fingers, the thumb and ring finger being curved inwards. Or, it is possible to avoid bringing on a person by speaking in the air (ākāçabhāṣita), pretending to hear the reply, and repeating it before answering it, while a similar purpose can be served by a voice from behind the scene.

The number of Acts which a play should contain varies according to the nature of the drama; in the Nāṭaka the number must be at least five, and may be ten; in other cases one Act suffices. Normally the Acts are simply numbered; in some cases, as in that of the Mṛcchakaṭikā, names are given, doubtless not by the poet.

4. THE CHARACTERS

The hero owes his name, Nāyaka, to the fact that it is he who leads (nī) the events to the conclusion which he has set before him, in so far as such a result is permitted by human frailty and the force of circumstances. His good qualities are innumerable [639]; he must be modest as is Rāma in depreciating his own prowess in comparison with that of Paraçurāma whom he has vanquished; handsome, generous like Jīmūtavāhana, prompt and skilled in action, affable, beloved of his people, of high family, ready of speech, and steadfast. He must be young, and endowed with intelligence, energy, a good memory, skill in the arts, and just pride; a hero, firm, glorious, skilled in the sciences, and an observer of law. More useful is the distinction drawn between types of hero [640]; all are noble or self-controlled (dhīra), a characteristic not universally found in heroines, but they are distinguished as light-hearted or gay (lalita), calm (çānta) exalted (udātta), and haughty or vehement (uddhata).

The light-hearted hero is one free from care, a lover of the arts, and above all a devotee of love; he is normally a king whose public burdens are confided to others, and whose one business it is to secure union with a new favourite by overcoming the obstacles interposed by the not unnatural jealousy of his queen or queens; such beyond all is Vatsa in Bhāsa and Harṣa’s dramas. The calm hero differs primarily from the light-hearted hero by reason of his birth, for he is a Brahmin or merchant, such as Mādhava in the Mālatīmādhava and Cārudatta in the Daridracārudatta and the Mṛcchakaṭikā; the hero of the Prakaraṇa, or comedy of manners, normally is of this class. The exalted hero is a character of great strength and nobility, firm of purpose, but free from vanity, forbearing, and without egotism. Of such a type are generals, ministers and high officials, and Jīmūtavāhana in the Nāgānanda. An instructive controversy rages round this description of Jīmūtavāhana; to be exalted, it is argued, implies the desire of superiority, but Jīmūtavāhana renounces every dream of empire and is a model of calm, of boundless pity, and freedom from passion save, indeed, as regards his love for Malayavatī, which is inconsistent with the general nature of his character. He should really be ranked among the calm heroes, with the Buddha himself, disregarding the meaningless convention which excludes kings from that category. Dhanika [641] effectively defends the classification of Jīmūtavāhana by insisting that he is not without desire, namely that of saving others at the cost of his own life; the desires he lays aside are wishes for personal advantage which Kālidāsa rightly censures in a king; his love for Malayavatī is wholly inconsistent with calmness, which, on the contrary, is in fact as in drama a characteristic of Brahmins, and it distinguishes him absolutely from the Buddha, who is exempt from passion. The haughty hero is a victim of pride and jealousy, an adept in magic arts and ruses, self-assertive, fickle, irascible, and boastful; Paraçurāma illustrates this character.

The chief hero in any drama must be essentially true to one or other of those types; any change would spoil the unity of the development of the drama, and, if necessary, changes must be made in the plot, as in the case of Rāma’s dealings with Vālin, to preserve the unity of character. In the case of the secondary hero there is no need for such consistency; he may change in different situations, and his lack of consistency tends merely to heighten the impression caused by the constancy of the hero. Thus Paraçurāma appears in the Mahāvīracarita [642] as exalted in his attitude to the evil Rāvaṇa, as haughty towards the untried Rāma, and as calm when he has experienced the superior prowess of that hero. It is obvious that there is difficulty in conceiving as a chief hero one of the haughty type, and the theory does not provide us with one, for Paraçurāma is only a secondary hero.

As the Sanskrit drama deals usually with love, the theory has another division of types of hero based on their attitude to women. [643] The courteous (dakṣiṇa) hero is one who can find room in his heart for more loves than one; he seeks another to the deep grief of the old, but he does not cease to feel affection for his earlier love; such are the heroes of the Nāṭikā, or short heroic comedy, like Vatsa. He may not be regarded as either deceitful (çaṭha), or shameless (dhṛṣṭa), for these two types represent heroes who have ceased to love their former flame, and differ only in so far as they seek to deceive her, or are indifferent to her anger and bear open traces of the new attachment. Men like Vatsa never allow passion to dominate them; if a woman spurns them they are ready to leave her. The fourth type is the loyal (anukūla) lover who is faithful to one woman only, as is Rāma. As these four types are applicable to each class of hero, there are sixteen possible kinds of hero, and the theory adds the further complication that each of these may be a high class, middle class, or inferior person, giving forty-eight types.

As if the enumeration of the general characteristics of the hero were insufficient, a set of eight special excellencies [644] is enumerated separately as springing from his character (sāttvika). These are brilliance (çobhā), including compassion for inferiors, emulation with superiors, heroism, and cleverness; vivacity (vilāsa), including a firm step and glance and a laughing voice; grace (mādhurya) manifested in the display of but slight change of demeanour in trying circumstances; impassivity (gāmbhīrya) or superiority to emotion; steadfastness (sthairya) in accomplishing his object despite obstacles; the sense of honour (tejas) which will punish insult even at the cost of life itself; lightheartedness as grace of deportment; and nobility (audārya) exhibited in sacrifice for the sake of the good.

The enemy of the hero (pratināyaka) [645] is self-controlled and vehement (dhīroddhata); but he is also avaricious, stubborn, criminal, and vicious; such are Rāvaṇa and Duryodhana as contrasted with Rāma and Yudhiṣṭhira. On the other hand, the hero of the episode, the companion (pīṭhamarda) [646] of the hero, is to possess, but in a less degree, the qualities of the hero; he is to be intelligent, ever in attendance on the hero, and devoted to his interests, as are Makaranda in the Mālatīmādhava and Sugrīva in the dramas based on the Rāma legend. The term, however, is unknown to these plays, while in the Mālavikāgnimitra the nun Kauçikī is styled a Pīṭhamardikā, and serves as a trusted go-between. The theory here seems to have stereotyped a relationship commoner in an older type of drama.

The heroine, Nāyikā, [647] plays a part in the economy of the drama similar to that of the hero, and not of less importance. The types of heroine depend primarily on her relation to the hero; she may be his wife (svā, svīyā) or belong to another (anyā, anyastrī) or be a hetaera. The hero’s wife must be upright and of good character, but she may be inexperienced (mugdhā), partly experienced (madhyā), or fully experienced and bold (pragalbhā). The inexperienced wife is shy in her love and gentle in her anger with her spouse’s infidelities. The partly experienced is full of the love of youth, and even faints in her passion; when angry, if self-controlled, she chides her husband with double entendres; if but partly controlled, she allows her tears to aid her reproaches; if uncontrolled, she adds harsh words. The bold wife is frantically in love, fainting at the first embrace; when angry, if self-controlled, she adopts an attitude of haughty reserve and indifference to the pleasures of life; if lacking in self-control, she uses threats and blows; if partly self-controlled, she employs the weapons of raillery and equivoke. A further division is possible, for each of these three kinds of heroine may be subdivided according as the lady is the earlier or later of the loves of the husband.

A woman, who is in the power of another, may be the wife of another man or a maiden. An amour with a married woman may not form the subject of the dominant sentiment in the play, but that for a maiden may occur as an element in the principal or the secondary action. Even when a parent or guardian is willing to permit a maiden’s marriage, there may be other obstacles, as in the case of the love of Mālatī and Mādhava and in Vatsa’s numerous amours. The woman who is common to all (sādhāraṇī) is a courtesan, skilled in the arts, bold, and cunning; she accepts as lovers the rich, the foolish, the self-willed, the selfish, and the impotent so long as their money lasts, then she has them turned out of doors by her mother, who acts as go-between. If she is a heroine, she must be represented as in love, like Vasantasenā in the Mṛcchakaṭikā, except in a Prahasana or farce, where she can be depicted as fleecing her lovers for comic effect; she must not figure as a heroine if the hero is divine or royal.

The heroine may occupy eight different relations to her lover. [648] She may be his absolute mistress (svādhīnapatikā) and he her obedient slave; she may be awaiting him in full dress (vāsakasajjā); she may be distressed by his involuntary absence (virahotkaṇṭhitā), enraged (khaṇḍitā) at discovering him disfigured by the marks of her rival’s teeth and nails, or be severed from her beloved by a quarrel (kalahāntaritā) and suffer remorse, or be deceived (vipralabdhā) by a lover who fails to meet her at the rendezvous which she has named. Her lover may be absent abroad (proṣitapriyā), or she may have to seek him out or press him to come to her (abhisārikā), giving as meeting-place a ruined temple, a garden, the house of a go-between, a cemetery, the bank of a stream, or in general any dark place. The first two classes of heroine are bright and gay, the others are wearied, tearful, changing colour, sighing, and wear no ornaments as token of their dejection. A woman, who is subject to another, cannot stand in all these relations to a lover; she may be distressed at his absence, deceived, or driven to seek him out, but she cannot be enraged, for she is not the mistress of her lover, and thus the king’s courtesy to Mālavikā in Kālidāsa’s play is not to be treated as an effort to appease an enraged heroine.

The heroine is accorded even a more generous allowance of excellencies than the hero. [649] The first three are physical, the first display of emotion in a nature previously exempt (bhāva), the movement of eyes and brows betokening the awakening of love (hāva), and the still more open manifestation of affection. The next seven are inherent characteristics of the heroine; the brilliance of youth and passion; the added touch of loveliness given by love, sweetness, radiance, courage, dignity, and self-control. Then come ten graces; the sportive imitation of the movements or words of the beloved one, the swift change of aspect at his arrival, tasteful arrangement of one’s ornaments to increase radiance of appearance, studied confusion of ornaments, hysteria (kilakiñcita), in which anger, fear, joy, and tears mingle, manifestations of affection (moṭṭāyita) on hearing the beloved mentioned or seeing his portrait, pretended anger (kuṭṭamita) on the lover touching hair or lip, affected indifference (bibboka), born of excess of pride, a graceful pose (lalita), and the bashfulness which forbids speech even when an opportunity presents itself. To these twenty Viçvanātha adds eight more graces; the pride which is vain of youth and beauty, the ennui which besets the maiden in her lover’s absence, the naïveté which displays itself in pretended ignorance and innocence, the distraction evinced by ornaments in disorder, wandering glances, and truant words, curiosity, the meaningless laugh of youth and high spirits, the tremors of fear causeless but common in the presence of the lover, and the sportive play of young affection. The same source gives us in great detail the modes in which the different types of heroine display their affection, in maidenly modesty or in shameless boldness, an analysis showing keen and deep insight into all the outward manifestations of love at an Indian court. Less praiseworthy is the perverse ingenuity which enumerates the different types of heroine, and educes first 128 from the combination of the eight forms of relationship to the lover with the sixteen kinds based on the division of wife, another’s, and hetaera. These are then multiplied by three on the basis of the division of all characters as high class, middle class, and low class.

The same division of classes is applied to all the other characters (pātra) which can appear in a play, but a much more fundamental classification is that by sex, masculine, feminine, and neuter. Most of the rôles are such as are incidental to the life of a palace, for the normal drama deals with the amours of a king, and his entourage and that of the queen account for practically all the normal characters of the drama.

The king’s confidant and devoted friend is the Vidūṣaka, [650] a Brahmin, ludicrous alike in dress, speech, and behaviour. He is a misshapen dwarf, baldheaded, with projecting teeth and red eyes, who makes himself ridiculous by his silly chatter in Prākrit, and his greed for food and presents of every kind. It is a regular part of the play for the other characters to make fun of him, but he is always by the king’s side, and the latter makes him his confidant in all his affairs of the heart, while the Vidūṣaka repays him by willing, if frequently incompetent or unlucky, attempts at service. The theorists offer no explanation of the anomaly of a Brahmin in such a curious position, but Açvaghoṣa already has the figure, as has Bhāsa, though not in his epic dramas, and later he is established as almost an essential feature in all dramas not derived from the epic; the chief exception is the Mālatīmādhava, where, however, his place is taken by the hero’s friend in sport (narmasuhṛd).

A much less common, but an interesting character is that of the Viṭa, [651] who resembles, though distantly, the parasite of the Greek drama; he is a poet skilled in the arts, especially music, acquainted au fond with the ways of hetaerae, in short a perfect man of the world with literary and artistic culture to boot. He is an essential figure in the Bhāṇa, or monologue, in which he relates his own shady adventures, but in other forms of drama he plays but a small part; Kālidāsa and Bhavabhūti ignore him, and, while Harṣa depicts him in the Nāgānanda, his position there is episodic; in the Mṛcchakaṭikā alone does he attain full development in his relation to the boastful Çakāra. Both these figures appear also in the Cārudatta, Çūdraka’s model. The Çakāra, [652] brother of a royal concubine, is of low caste, easily angered and appeased, fond of fine raiment, and proud of his office, in which, however, he shows himself corrupt and incompetent. He is found also in an episode of the Çakuntalā, but then fades from the drama leaving, however, a clear suggestion of its early history.

The king requires in his amours the aid of a messenger (dūta) [653] as well as for more serious affairs. The holder of this rôle must be possessed of loyalty, energy, courage, a good memory, and adroitness; he may be given full powers to act as seems best in each emergency, or have limited authority, or be a mere bearer of a message. Others intimately associated with the royal household are the servants (ceṭa), [654] the mercenaries, Kirātas or Mlecchas, the chaplain, priest, and other theologians. There are also those employed in the government of the realm, which the king is only too pleased to neglect. [655] The minister (mantrin, amātya) is of good family, of high intelligence, skilled in affairs human and divine, and devoted to the interests of the country. The general (senāpati) is also of high birth, incapable of weakness, skilled in both the theory and the practice of war, and kind of speech; ready to note the weakness of the enemy and to direct at the suitable moment a campaign against him. The judge (prāḍvivāka) must be master of the laws and of judicial procedure, absolutely impartial, devoted to his duty, free from anger or pride, master of himself. The other officers are required to possess high qualities of intelligence, activity, and devotion to duty, while for less important work the king commands the services of foresters, military officers, and soldiers. The prince royal (kumāra) and the friend are also mentioned in the Nāṭyaçāstra, but without detail.

Of women’s rôles [656] the most important in dignity is that of the chief queen (mahādevī), the equal in age and rank of her husband, whose lapses in affection wound her, without robbing her of her sense of self-respect or dignity. In good fortune or evil she is devoted to him and seeks ever his welfare. The queen (devī) is also a daughter of a king, but she is more proud than dignified, and, intoxicated by her youth and beauty, her mind is set on the pleasures of love. The favourite (svāminī) is the daughter of a general or a minister, seductive by her beauty and intelligence, honoured by the king and others. There are other types of concubine (sthāyinī and bhoginī) with characteristics but little distinctive. The harem includes also the chief attendants (āyukta), who are charged with the supreme oversight of some department of the court, the king’s personal attendant who is always with him (anucārikā), the maid who performs his toilet and holds over him the umbrella of state, the women—called sometimes Yavanīs, once Greek maidens—who act as his body-guard, and those aged women who are skilled in political traditions and are respected on that score. There are also the princess, ingénue and modest, and the duenna (mahattarā), who among other things sees to the punctual performance of auspicious rites, and the more humble adepts in the dance, in song, in handicrafts, in acting, and in the favourite amusement of swinging the ladies of the harem. The hetaera is painted in attractive colours; she is thoroughly well educated, exempt from the normal defects of women, kind of heart, adroit, active, a born coquette, and seductive in every way. Special importance among these feminine rôles attaches to that of the heroine’s messenger, the counterpart of the hero’s agent. She may be a friend, a slave, a foster-sister, a neighbour, a workwoman, or an artiste, or strangely enough, a nun, usually of Buddhist connexions, a curious and interesting sidelight on Indian views of the devotees of that faith. The doorkeeper (pratīhārī) has the function of announcing to the king such political events as the declaration of war and the conclusion of peace.

The neuter rôles [657] are filled by men who have either taken vows of chastity, or have been deprived of virility in order to permit of their employment in the harem. The Snātaka is a Brahmin, who has completed his course of religious study, is familiar with religious and social affairs; he resides within the palace. The chamberlain (kañcukin) is an old Brahmin, worn out in the service of the king, but still mentally alert and skilled in his business of conveying the royal orders in the palace. The eunuchs (varṣadhara, nirmuṇḍa, upasthāyika) are effeminate and cowardly but not lacking in savoir faire; they find employment in the king’s amours.

The nomenclature [658] of the characters is in some measure regulated by rule; the name of a hetaera should end in dattā, senā, or siddhā, as does that of Vasantasenā in the Cārudatta; that of a merchant in datta as in Cārudatta; that of the Vidūṣaka from spring or a flower, but in the Avimāraka he is styled Saṁtuṣṭa; that of a servant, male or female, should be derived from some object, which occurs in descriptions of the seasons, &c., as in the names Kalahaṅsa and Mandārikā in the Mālatīmādhava; those of Kāpālikas, a species of ascetics, should end in ghaṇṭa as in Aghoraghaṇṭa in the same play.

There is also an elaborate etiquette [659] as to the mode of addressing the diverse personages. A king is styled thus by ascetics, but Deva or Svāmin by his courtiers; his charioteer and Brahmins generally hail him as Āyuṣmant, ‘long-lived’, while inferiors style him Bhaṭṭa, ‘master’. The crown prince is styled Svāmin, like his father; the other princes of the blood (bhartṛdāraka), but also common people, Bhadramukha or Saumya, preceded by he in the latter case, terms designed to conciliate by attributing to those addressed the qualities they are desired to show. [660] The style Bhagavant, ‘blessed’, is appropriate to the gods, to great sages and saints; Ārya, ‘noble’, is appropriate to Brahmins, ministers, and elder brothers, while a wife should address her husband as Āryaputra. Sages address an ascetic as Sādhu; ministers are styled Amātya or Saciva; the king calls his Vidūṣaka, and is called by him, Vayasya, ‘friend’. Sugṛhītābhidha, ‘well named one’, is the address [661] of a pupil to his master, a son to a father, or a younger to an elder brother, while the latter in return uses Tāta or Vatsa, both affectionate and condescending terms, suitable also for use to a son, or any person who owes one respect. Heretics should be given the styles they affect, thus a Buddhist should be hailed as Bhadanta; Çakas should be styled by such terms as Bhadradatta. The interjection Haṅho may be used between men of middle rank, Haṇḍe between common people. The Vidūṣaka addresses the queen and her ladies as Bhavatī; otherwise the queen is styled Bhaṭṭinī or Svāminī, a wife as Āryā, a princess Bhartṛdārikā, a hetaera Ajjukā, a go-between or aged dame Ambā; Halā is used between friends of equal rank, Hañjā is addressed to a servant.

5. THE SENTIMENTS

The most original and interesting part of dramatic theory is the gradual definition of the nature of the sentiment which it is the aim of the performance to evoke in the mind of the audience. [662] The statement of the Nāṭyaçāstra is simple. Sentiment is produced from the union of the determinants (vibhāva), the consequents (anubhāva), and the transitory feelings (vyabhicārin). The determinants fall in the later classification into two divisions, the fundamental determinants (ālambana) and the excitant determinants (uddīpana); fundamental determinants comprise such things as the heroine or the hero, for without them there can be no creation of sentiment in the audience; excitant determinants are such conditions of place and time and circumstance as serve to foster sentiment when it has arisen, for instance the moon, the cry of the cuckoo, the soft breeze from Malaya, all things which foster the erotic sentiment. The consequents are the external manifestations of feeling, by which the actors exhibit to the audience the minds and hearts of the persons of the drama, such as sidelong glances, a smile, a movement of the arm, and—though this is but slightly indicated in later texts—his words. [663] A special class is later made of those consequents, which are the involuntary product of sympathetic realization of the feeling of the person portrayed, and hence are called Sāttvika, as arising from a heart which is ready to appreciate the sorrows or joys of another (sattva); these are paralysis, fainting, horripilation, perspiration, change of colour, trembling, weeping, and change of voice. The transitory or evanescent feelings are given as thirty-three; they are discouragement, weakness, apprehension, weariness, contentment, stupor, joy, depression, cruelty, anxiety, fright, envy, indignation, arrogance, recollection, death, intoxication, dreaming, sleeping, awakening, shame, epilepsy, distraction, assurance, indolence, agitation, deliberation, dissimulation, sickness, insanity, despair, impatience, and inconstancy. But these factors are not sufficient to account for sentiment, nor does the Nāṭyaçāstra intend this. It recognizes that an essential element in the production of sentiment is the dominant emotion (sthāyibhāva) which persists throughout the drama amid the variations of the transitory feelings; it stands to the other factors in the position of the king to his subjects or a master to his pupils, as the Çāstra says; it is, says the Daçarūpa, the source of delight, and brings into harmony with itself the transitory states of feeling.

It is the dominant emotions which in some fashion determine or become sentiments even in the view of the Nāṭyaçāstra, though in it there is clearly difficulty in conceiving the precise signification of the process, a fact revealed in a tendency to confuse the terms emotion and sentiment, Bhāva and Rasa. In Bhaṭṭa Lollaṭa [664] we have a determined effort to make clear the implication of the doctrine. The dominant emotion of love, for instance, generated by a fundamental determinant such as a maiden, inflamed by an excitant determinant such as a pleasant garden, made cognizable by consequents such as sidelong glances and embraces, and strengthened by transitory feelings such as desire, becomes the erotic sentiment first of all in the hero of the drama, e.g. Rāma. The sentiment is subsequently attributed to the actor who imitates the hero in form, dress, and action, and so it becomes the source of charm to the audience. The fatal objection to this theory is clear; it fails to recognize that the sentiment must be that of the spectator himself; he cannot have enjoyment of a sentiment which exists merely in the actor as a secondary outcome of its existence in Rāma. Moreover, the actor whose chief aim is to please the audience and earn money need not feel at all the emotions of Rāma, while, if he does so, he is then in the same position as a spectator.

The view of Lollaṭa, which is classed as one of the production (utpatti) of sentiment and regarded as that of the Mīmāṅsā school, is opposed by the doctrine of Çrīçan̄kuka, regarded as the Naiyāyika view, which interprets the manifestation of sentiment as a process of inference. The emotions, love, &c., are inferred to exist in the actor, though not really present in him, by means of the determinants, &c., cleverly exhibited in his acting; the emotion thus inferred, being sensed by the audience, through its exquisite beauty, adds to itself a peculiar charm and thus finally develops into the state of a sentiment in the spectator. This view, however, is open to the fatal objection that it is commonly admitted that it is not inference, or any other derivative mode of knowledge, which produces charm, but perception alone, and no adequate ground exists for disregarding this general truth in this case.

In Bhaṭṭa Nāyaka [665] we find yet a different point of view, which denies either the production (utpatti) of sentiment, its perception or apprehension (pratīti) or its revelation (abhivyakti). If sentiment is perceived as appertaining to another, then it cannot personally affect oneself. But it cannot be perceived as present in oneself as a result of study of a work about Rāma; there are no factors present in the self to produce any such result; it is impossible to hold that an emotion dormant in oneself is called to life by seeing or reading the story of Rāma; experience shows that one’s own beloved does not come up to one’s mind to raise love, nor could a tale of a goddess evoke the picture of a mortal amour; again, such marvellous deeds as Rāma’s have nothing common to mortal efforts so as to be able to awake conceptions of acts of our own. Thus sentiment cannot be apprehended. Nor is it a case of production; if so, no one would go twice to a play of a pathetic type, since one would experience actual misery as the result, in lieu of a pleasant melancholy; again the sight of lovers united does not in real life produce sentiment. Nor is a case of the revelation of something existing potentially (çaktirūpa). If this were so, then, when the potential emotions were let loose, they would occupy their field of action in diverse degrees—thus contradicting the nature of sentiment as one; moreover there would be the same difficulties as in the case of apprehension as to whether revelation applied to the hero or oneself. The true solution, therefore, is to ascribe to a poem a peculiar threefold potency of its own, the power of denotation (abhidhā), which deals with what is expressed, the power of realization (bhāvakatva), which relates to the sentiment, and the power of enjoyment (bhojakatva), which has regard to the audience. If denotation were all, there would be no difference between poetic figures and manuals, there would be absence of the distinctions produced by divergence of literal and metaphorical sense, and the avoidance of harsh sounds would be needless. As it is, we have the second function of realization of sentiment, which causes the expressed sense to serve as the basis of the sentiment, and confers on the determinants, &c., the essential feature of being appropriated by the audience as universal. From this comes the appreciation by the audience of the sentiment, an appreciation consisting in a mental condition made up entirely of the element of goodness or truth (sattva), uninfluenced by the other elements of passion (rajas) and dullness (tamas), that is, entirely free from desire, comparable with meditation on the absolute. This condition is the vital element; the enjoyment ranks above the aesthetic equipment [666] which renders it possible. To this theory which is sometimes ascribed to the Sāṁkhya, [667] and called the Bhuktivāda, doctrine of the enjoyment of sentiment, the objection is made that the two powers ascribed to poetry, realization, and enjoyment, have no legitimate foundation.

The view finally adopted by the theorists is that defended, but not first enunciated, by Abhinavagupta, based on the general doctrine of suggestion (vyañjanā) as lying at the basis of all poetic pleasure. The spectator’s state of mind must be considered; it is in him that from experience of life there come into being emotional complexes, which lie dormant, ready to be called into activity by the reading of poems or by seeing plays performed. Those whose life has left them barren of impressions of emotions are, accordingly, incapable of relishing dramas, a fate which awaits men whose minds are intent merely on grammar or on the complexities of the Mīmāṅsā. The sentiment thus excited is peculiar, in that it is essentially universal in character; it is common to all other trained spectators, and it has essentially no personal significance. A sentiment is thus something very different from an ordinary emotion; it is generic and disinterested, while an emotion is individual and immediately personal. An emotion again may be pleasant or painful, but a sentiment is marked by that impersonal joy, characteristic of the contemplation of the supreme being by the adept, a bliss which is absolutely without personal feeling. There is in fact a close parallel between the man of taste (sahṛdaya) [668] and the adept (yogin); both have in them the possibility of attaining this bliss, and, to make it real, the one must investigate the determinants, &c., while the other must apply himself to concentration on the absolute. It is this peculiar nature of sentiment which forbids it being created as the result of denotation or indication by speech, of perception, inference, or recollection. It cannot exist without determinants, &c., but these are not in the normal sense causes; an effect can exist when its causes have disappeared, but sentiment exists only while the determinants, &c., last; the terms used in this regard are one and all distinct from the normal terminology of causation. Sentiment is something supernatural (alaukika); its relations to the factors may be compared with that of a beverage to the black pepper, candied sugar, camphor, &c., which compose it, but of which as such no trace remains in the liquor as produced. This characteristic enables us to understand how it is that the list of sentiments includes that of horror or odium (bībhatsa) and that of fear (bhayānaka), as well as the pathetic sentiment. These are awakened into life by things which cause disgust, fear, and grief in ordinary life, and these emotions in real life are far from pleasant in any sense of that term. But, conveyed as ideal and generic, they produce this supernatural pleasure or happiness, which is not to be compared with normal pleasure, just as the joy of the contemplation of the absolute is not to be described as pleasure in the ordinary sense. Bhānudatta, in his Rasataran̄giṇī, a work composed before A.D. 1437, distinguishes Rāsa as natural (laukika) and supernatural or transcendental; the former is the emotion experienced in ordinary life—which may more conveniently be distinguished as Bhāva,—the latter includes the emotion experienced in dream experiences, in the building of castles in the air, and in the appreciation of poetry, and he is careful to emphasize the totally different nature of the natural and the transcendental emotion. [669]

The doctrine set out in Abhinavagupta is also that of the Daçarūpa, although it is rendered more obscure there by the brevity of its exposition. The process of transformation of an emotion to a sentiment is formally described; ‘a dominant feeling or emotion becomes a sentiment when it is transformed into an object of enjoyment through the co-operation of the determinants, the consequents, including the involuntary manifestations of feeling, and the transitory feelings’. [670] The sense is made further precise by the assertion [671] that the dominant emotion becomes a sentiment, because it is enjoyed by the spectator of taste, and he is actually at present in existence; the sentiment is not located in the hero whose actions are represented, for he belongs to the past, nor does it appertain to the poem, for that is not the object of the poem—its function being to set out the determinants, &c., through which the dominant emotion is brought out and generates the sentiment,—nor is sentiment the apprehension by the spectator of the emotions enacted by the actor, since in that case spectators would feel not sentiment, but an emotion varying in the different individuals, just as in real life from seeing a pair in union those who see them feel according to their nature shame, envy, desire, or aversion. The position of the spectator is compared to that of the child which, when it plays with its clay elephants—the ancient equivalent of our tin soldiers—experiences the sensation of its own energy as pleasant; the deeds of Arjuna arouse a like feeling in the spectator’s mind. This experiencing sentiment is a manifestation of that joy which is innate as the true nature of the self, and this manifestation comes into being as the result of the pervasion of the mind of the spectator with the dominant emotion and the determinants, &c., in combination.

An effort is made to describe the precise nature of the mental activity involved in the enjoyment of sentiment, and to base upon it a division of the sentiments. The four sentiments of love, heroism, horror, and fury are taken as primary, and brought into connexion with mental conditions described as the unfolding (vikāsa), expansion (vistara), agitation (kṣobha), and movement to and fro (vikṣepa) of the mind. [672] These are evidently mental conditions, believed to be reached by introspection, and they have the merit of giving a quasi-psychological rationale for the doctrine of four primary and four secondary sentiments found in the Nāṭyaçāstra. [673] But there was no early agreement on this piece of psychology; Abhinavagupta, [674] with Bhaṭṭa Nāyaka, accepts only three aspects of mental condition as involved, the melting (druti), expansion, and unfolding, a division which is applied also in the theory of poetics to justify the doctrine of the existence of three qualities only of words. [675] On Dhanaṁjaya’s view the sentiment of calm which he denies for drama, [676] if it exists at all, must be regarded as combining all the four mental aspects above distinguished.

It is now possible to understand clearly the essential relation of the spectator to the actors; we see on the stage, for instance, Rāma and Sītā, who excites his affection, aided by suitable circumstances of time and place; this affection is intimated by speech and gesture alike, which indicate both the dominant emotion of love and its transient shapes in the various stages of love requited. The spectacle evokes in the mind of the spectator the impressions of the emotion of love which experience has planted there, and this ideal and generic excitation of the emotion produces in him that sense of joy which is known as sentiment. The fullness of the enjoyment depends essentially on the nature and experience of the spectator, to whom it falls to identify himself with the hero or other character, and thus to experience in ideal form his emotions and feelings. He may even succeed in his effort to the extent that he weeps real tears, feels terror and sorrow, but the sentiment is still one of exquisite joy. We may compare the thrill of pleasure which the most terrifying narration excites in us, and we are all conscious of the sweetness of sad tales.

Viçvanātha insists very strongly on the necessity of the identification of the spectator with the personages depicted, a process which enables him to accept without any difficulty such episodes of extraordinary character as Hanumant’s leap over the ocean. [677] He must not treat the emotion of love as his own, for in that case it would never become a sentiment; it would remain a feeling, and in the case of fear, for instance, it would cause pain, not joy. Nor must he regard it as belonging solely to the hero, for then it would remain his feeling, and in no wise affect the spectator or become a sentiment. Similarly, the determinants, &c., are not to be treated as pertaining to the hero alone; they must be felt as generic. This generic action (sādhāraṇī kṛti) is the essential feature, replacing the generic power which Bhaṭṭa Nāyaka attributes to poetry. We can now see clearly the position of the actor; the Nāṭyaçāstra [678] bids him as far as possible to assume the emotions of the person whom he represents, and to depict them by costume, speech, movements, and gestures as his own, but Viçvanātha [679] is more anxious to insist that the sentiment is not necessarily to be found in the actor, who often merely performs mechanically his part according to rote and rule; if he actually does experience the feelings he portrays, then he becomes in so far a spectator. [680] Further, he points out the simultaneous presence of all the factors is by no means essential, for the existence of one will revive the others by force of the association of ideas. He insists also on the necessity of experience and cultivation of the power of imagination in one who seeks to enjoy sentiment; as we are by virtue of the doctrine of transmigration—or if we prefer to modernize, by heredity—endowed with the germs of the capacity of appreciation, we can normally by study of poetical works develop the capacity, but, if we devote ourselves to the study of grammar or philosophy, we shall certainly deaden our susceptibilities. The difficult problem, why much study of poetry leaves some still unable to relish the sentiment, is explained by the convenient hypothesis that demerit in a previous birth intervenes to frustrate present effort. He refutes at length the effort of Mahiman Bhaṭṭa [681] to destroy the whole doctrine of suggestion in poetry by the doctrine of inference; doubtless by inference we could arrive at a belief in the existence of an emotion in the hero’s mind but that inference would not produce any effect in us or arouse sentiment; a logician might make the inference and draw the correct conclusion, but would remain cold and unmoved. Suggestiveness, he shows, is absolutely essential as a function of words and as the characteristic of poetry, giving it power to create sentiment. What is expressed may be understood by every one; the man of taste alone appreciates the suggestion and enjoys the flavour resulting.

Now sentiment is one, it is a single, ineffable, transcendental joy, but it can be subdivided, not according to its own nature, but according to the emotions which evoke it. Thus the Nāṭyaçāstra recognizes the existence of eight emotions or dominant feelings; love (rati), mirth (hāsa), anger (krodha), sorrow (çoka), energy (utsāha), terror (bhaya), disgust (jugupsā), and astonishment (vismaya), and corresponding to these eight emotions we have eight forms of sentiment. The erotic sentiment (çṛn̄gārarasa) is of two kinds, the union (sambhoga) or sundering (vipralambha) of two lovers, according to the Çāstra and the great mass of theorists, but the Daçarūpa [682] distinguishes three cases, privation (ayoga), sundering (viprayoga), and union. Privation denotes the inability of two young hearts to secure union, because of obstacles to their marriage; such love passes through ten stages, [683] longing, anxiety, recollection, enumeration of the loved one’s merits, distress, raving, insanity, fever, stupor, and death. Sundering may be due to absence or resentment, and this in its turn may be caused by a quarrel between two determined lovers, or indignation at finding out, by sight, hearing or inference, that one’s lover is devoted to another. The hero may counteract anger by conciliation, by winning over her friends, by gifts, by humility, by indifference, and by distracting her attention. Absence again may be due to business, to accident, or a curse; if the reason is death the love sentiment cannot, in Dhanaṁjaya’s view, be present, though others allow of a pathetic variety of this sentiment. [684] In union the lover should avoid vulgarity or annoyance.

The heroic (vīra) sentiment corresponds to the emotion of energy; it may take the three forms of courage in battle as in Rāma; compassion as in Jīmūtavāhana; and liberality as in Paraçurāma. Assurance, contentment, arrogance, and joy are the transitory states connected with it. The sentiment of fury (raudra) is based on anger; its transitory states are indignation, intoxication, recollection, inconstancy, envy, cruelty, agitation, and the like. The comic (hāsya) sentiment depends on mirth, which is caused by one’s own or another’s strange appearance, speech, or attire. [685] The transitory states in connexion with it are sleeping, indolence, weariness, weakness, and stupor. The sentiment of wonder (adbhuta) is based on astonishment; the transitory states are usually joy, agitation, and contentment. The sentiment of terror (bhayānaka) is based on terror; the states associated with it are depression, agitation, distraction, fright, and the like. The pathetic (karuṇa) sentiment is based on sorrow; its associated states are sleeping, epilepsy, depression, sickness, death, indolence, agitation, despair, stupor, insanity, anxiety, and so forth. The sentiment of horror or odium (bībhatsa) is based on disgust; its associated states are agitation, sickness, apprehension, and the like. In each case the theorists give in full the determinants and the consequents of each emotion, which becomes a sentiment, and a special colour is ascribed to each; it is not surprising to find that red is associated with fury, black with fear; whiteness may, in association with the comic sentiment, be explained by the flashing teeth of a laughing maiden, and the dark (çyāma) colour of the erotic sentiment is a reflex of the favoured hue of the beloved; grey accords with pathos, but the connexion of yellow with wonder, dark blue with horror, and orange with heroism is not obvious. It is also artificial to find four primary and four secondary sentiments laid down; the erotic, the furious, the heroic, and that of horror, whence in order are supposed to develop the comic, the pathetic, that of wonder and that of terror. The Nāṭyaçāstra recognizes these eight only, [686] but later authorities add the sentiment of calm (çānta) based on indifference to worldly things (nirveda), although this is in the Çāstra merely a transitory feeling. Those who follow the Çāstra contend that there is no such sentiment, for it is impossible to destroy utterly love, hatred, and other feelings, which have been operative from time without beginning; others admit the existence of the sentiment, as does Mammaṭa, but not in drama, on the ground that indifference to all worldly things is incapable of being represented. But this also is erroneous; the actor’s power of representing indifference is not in point, as it is the spectator who is to feel the sentiment, and the fact that the Çāstra places it first in the list of transitory states, though that would normally be an inauspicious beginning, indicates that it was meant to serve both as an emotion and a transitory feeling, and it is fully recognized by Vidyādhara, Viçvanātha, and Jagannātha, though Dhanaṁjaya barely admits it. [687] The interrelations of the sentiments, their possible combinations, their harmonies and conflicts, are detailed at length.

The sentiments may all be employed in drama, but there are rules affecting their use. In each play there should be a dominant sentiment; in the Nāṭaka it should be the erotic or the heroic; other sentiments are merely auxiliary, but that of wonder is especially appropriate in the dénouement; indeed something in the way of supernatural intervention is often convenient to extricate the plot. An excess of sentiments is as bad as a defect; if there are too many they destroy the unity of the whole and detach it into a series of ill-connected fragments, while the excessive use of action and of rhetorical display is equally destructive to the merit of a piece.

The Çakuntalā illustrates excellently the sentiment of love as the ruling motive of the play; the heroic sentiment appears in the verses in Act II in which the hermits extol the king; the horrible in Act VI in the scene in which Mātali menaces the Vidūṣaka; terror is evoked by the description of the dusk at the close of Act III; the whole play from the arrival of Kaṇva in Act IV to the departure of Çakuntalā produces the sentiment of pathos, while that of fury is called into being by the close of Act VI from the despairing cries of the Vidūṣaka to the entry of Mātali; finally wonder is aroused by the strange incident at the close when the king picks up the bracelet fallen from the arm of the child which, unknown to him, is his own son by the wife whom he has in ignorance repudiated. The Nāṭikās afford excellent examples of the erotic sentiment; Harṣa, in complete accord with the rules of the drama, helps out his plot in both the Ratnāvalī and the Priyadarçikā by the use of incidents evoking the sentiment of wonder; the imprisonment of Sāgarikā in the former play evokes the sentiment of pathos, while terror is excited by the description in Act II of the wild confusion caused by the monkey’s escape from the royal mews. The sentiment of fury is frequently evoked in the Mahāvīracarita and the Veṇīsaṁhāra; the Mālatīmādhava affords excellent illustrations evoking horror, while the Mahāvīracarita is permeated by the sentiment of heroism. The Nāgānanda reveals heroism in another aspect, that of the perfection of compassion and nobility, for, as we have seen, Jīmūtavāhana is not to be regarded as a hero in whom calm prevails.

There is doubtless pedantry in the theory of sentiment; the choice of eight emotions, the subordination to them of transitory states, the enumeration of determinants and consequents, are largely dominated by empiricism, and not explained or justified. But in its essentials the theory may be admitted to be a bold and by no means negligible attempt to indicate the essential character of the emotional effect of drama.

6. THE DRAMATIC STYLES AND LANGUAGES

Plot, characters, and sentiment are not the only constituent elements of drama; the poet must be an adept in adopting the appropriate manner [688] or style (vṛtti), for each action of the hero; the style adds to the play the indefinable element of perfection which is present in the highest beauty of feature or dress. The manners allowed by the Nāṭyaçāstra are four, the graceful (kaiçikī), the grand (sāttvatī), the violent (ārabhaṭī), and the verbal (bhāratī), which owes its name to the fact that, unlike the others, it depends for its effect on words, not action.

The graceful manner is appropriate to the erotic sentiment; it employs song, dance, and lovely raiment, admits both male and female rôles, and depicts love, gallantry, coquetry, and jesting. It admits of four varieties. The first is pleasantry (narman), which is based on what is comic in speech, dress, or movement in the actors; the pleasantry may be purely comic, or be mingled with love, or even with fear, as when Susaṁgatā makes fun of Sāgarikā and adds that she will tell the queen of the episode of the picture. [689] When love is mingled, it may serve to evince affection, or to ask for a response, or to impute a fault on the lover’s part. A comedy of costume is seen in the Nāgānanda where the Viṭa, misled by his garments, mistakes the Vidūṣaka for a woman; a comedy of action in the Mālavikāgnimitra where Nipunikā punished the Vidūṣaka by dropping on him a stick which he takes, naturally enough, for a snake. The second form is the outburst of affection (narmasphūrja) [690] at the first meeting of lovers, which ends in a note of fear, as in the meeting of the king and Mālavikā in Act IV of the Mālavikāgnimitra. Thirdly, there is the manifestation of a recent love by physical signs (narmasphoṭa), [691] and, fourth, the development of affection (narmagarbha), illustrated by the adoption of a disguise by the hero to attain his end, as when Vatsa in the Priyadarçikā comes on the scene in the garb of Manoramā. [692]

The grand manner is appropriate to the sentiments of heroism, wonder, and fury, and in a less degree to the pathetic and erotic. Virtue, courage, self-sacrifice, compassion, and righteousness are its subjects, not sorrow. Its divisions are the challenge (utthāpaka), as in the Mahāvīracarita, Act V, Vālin defies Rāma; breach of alliance (saṁghātya) among one’s foes, which may be brought about by deliberate stratagem, as in the Mudrārākṣasa, or by fate, as in the Rāma dramas Vibhīṣaṇa severs himself from Rāvaṇa; change of action (parivartaka) as when in the Mahāvīracarita Paraçurāma offers to embrace Rāma, whom he came to overthrow; and the dialogue (saṁlāpa) of warriors such as that of Rāma and Paraçurāma in the same play.

The violent manner accords with the sentiments of fury, horror, and terror. It employs magic, conjuration, conflicts, rage, fury, and underhand devices. Its elements include, first, the almost immediate construction (saṁkṣipti) of some object by artificial means, such as the elephant of mats made to contain Udayana’s men in the lost Udayanacarita; but others interpret this member as a sudden change of hero, whether real, as in the substitution of Vālin for Sugrīva, or merely a change of heart on the hero’s part, as in Paraçurāma’s submission to Rāma; in either case only a secondary hero can change or be changed, else the unity of the drama would disappear. The other elements are the creation of an object by magic means (vastūtthāpana); the angry meeting of two persons who end by fighting (sampheṭa), as do Mādhava and Aghoraghaṇṭa in the Mālatīmādhava; and a scene of tumultuous disturbance (avapāta), such as that when the monkey escapes in the Ratnāvalī or of the attack on Vindhyaketu in the Priyadarçikā, Act I.

The verbal manner is based on sound, as the other three are on sense. The voice only is its means of expression; women may not use it, and the men must speak Sanskrit; these actors bear the name Bharata, which is appropriated to this manner. It is adapted to all the sentiments, or, according to the Nāṭyaçāstra, only to those of heroism, wonder, and fury. Its elements are, in true scholastic fashion, likewise reckoned as four; two of them, the propitiation (prarocanā), and the introduction (āmukha, prastāvanā), essentially belong to the prologue of the drama, and will be considered in that connexion; the other two are given as the garland (vīthī) and the farce, which are species of drama. But the theorists agree that the elements (an̄ga) of the garland [693] are applicable in any part of the drama, especially the first juncture, and they are evidently an essential part of the verbal manner.

The first element is the abrupt dialogue (udghātya), which takes either the form of a series of questions and answers in explanation of something not at once understood, or a monologue of question and reply. The second is continuance (avalagita) of one section by another in substitution, as where, when Sītā has decided to go to the forest for pleasure, Rāma is persuaded to let her go indeed, but into exile, or, according to Dhanaṁjaya alone, where there is a sudden turn in an event in progress. [694] The third is the Prapañca, which passes for a comic dialogue, in which two actors frankly set out each other’s demerits, [695] or, according to Viçvanātha, such a clever ruse as that of Nipuṇikā in the Vikramorvaçī, Act II, where she worms out from the Vidūṣaka the king’s infatuation. The triple explanation (trigata), a term which is used in a different sense in the rule regarding the prologue, seems to denote guesses made at the cause of a sound, which in its character is ambiguous and may be, e.g. the hum of the bees, the cry of the cuckoo, or the music made by celestial maidens. [696] Cheating (chala) denotes the use of words of seeming courtesy but boding ill, as in the inquiry for Duryodhana, their foe, by Bhīma and Arjuna in the Veṇīsaṁhāra, Act V. The repartee (vākkelī) produces comic effect in a series of questions and answers; but the same term is applied to the interruption of a sentence by Dhanaṁjaya, and by Viçvanātha to a single reply to many questions. Outvying (adhibala or atibala) applies to a dialogue in which those conversing vie with one another in violence, as in the discussion of Arjuna, Bhīma, and Duryodhana in the Veṇīsaṁhāra, Act V. The abrupt remark (gaṇḍa) is one which intervenes vitally in the tale; thus in the Uttararāmacarita Rāma has just declared that separation from Sītā would be unbearable, when the porteress announces Durmukha, the spy of the king, who comes to destroy the king’s happiness. Reinterpretation (avasyandita) is the taking up of an expression which has escaped one in a different sense; thus in the Chalitarāma, Sītā carelessly tells her sons to go to Ayodhyā and greet their father, and seeks to remedy this slip by insisting that the king is father of his people. The enigma (nālikā) conceals the sense under joking words. Incoherent talk (asatpralāpa) is the speech of one just awake, drunk, asleep, or childish; such are the hero’s words in Vikramorvaçī, Act IV. In another sense, admitted by Viçvanātha, it denotes good advice thrown away, as in the Veṇīsaṁhāra, Act I, Gāndhārī’s admonition of Duryodhana. Humorous speech (vyāhāra) is a remark made for the sake of some one else, which provokes a laugh, as when the Vidūṣaka in the Mālavikāgnimitra, Act II, by his chatter makes the damsel laugh, and permits the king longer to gaze on her charms. Mildness (mṛdava) denotes the turning of evil into good, or vice versa, as when in the Çakuntalā, Act II, the virtues of hunting, a vice in the eyes of the sacred law, are extolled.

It is an essential defect of Indian theory in all its aspects that it tends to divisions which are needless and confusing. Besides the elements of the garland we find thirty-three dramatic ornaments (nāṭyālaṁkāra) [697] and thirty-six characteristics or beauties (lakṣaṇa), [698] which cannot be distinguished as two classes on any conceivable theory, [699] for both consist largely of modes of exposition and figures of thought and diction, while they also contain, as recognized by Dhanaṁjaya, a number of feelings which fall within the sphere of sentiment and its discussion. Thus we find as ornaments the benediction, the lamentation, raillery, the use of argument to support a view (upapatti), the prayer, the expression of resolution, the reproach, the provocation, the adduction of a common opinion in order to administer covertly a rebuke, the request, the narrative, reasoning, and the telling of a story. The beauties again include the combination of merits of style with poetic figures; the grouping of letters to make up a name; the use of analogy and example; the citation of admitted facts to refute incorrect views; the fitting of expression to the sense; the explanation by reasoning of a fact which is not capable of sense perception; the description of an object from the point of view of place, time, or shape; the indication of a characteristic which serves to distinguish two objects otherwise alike; the allusion to the truth of the literal meaning of a name; the use of the names of famous persons in a eulogy of some living being; the expression unconsciously, under the influence of passion, of the contrary of what one means; the statement in succession (mālā) of several means to attain a desired object; the expression of two different views, one of which in reality strengthens the other; the reproach; the question; the use of commonplaces; eulogy; the employment of a comparison to convey a sense which it is not desired directly to express; the indirect expression of desire; the veiled compliment; and the address of gratitude. Unfortunately no scientific attempt at orderly arrangement or examination of the principles on which these matters are based is attempted.

The Nāṭyaçāstra [700] adds an account of four ornaments of the drama (nāṭakālaṁkāra), which the Daçarūpa ignores, doubtless for the adequate reason that these matters appertain to poetics in general, and they are treated of in vast detail by the text-books of that science. The first is simile, defined as a comparison based on the similarity of characteristics in two objects; there are five kinds, the simile which extols, that which condemns, that with an imagined thing, as when the elephant is likened to a winged mountain, that based on similarity, and that on partial similarity, as in ‘Her face is like the full moon, her eyes like the blue lotus’. The metaphor is an abridged comparison which unites the two objects so as to efface their distinction, as ‘The fisher Love casts on the ocean of this world his lure, woman’. The illuminator (dīpaka) is the figure of speech, which uses but one verb to express the connexion between a subject series and a string of qualifications. Of forms of alliteration (yamaka), the repetition of vowels and consonants forming different words and meanings, there are enumerated ten kinds, a striking proof of the importance attached to these verbal jingles in early poetics.

The Çāstra [701] adds some vague and valueless suggestions as to the use of these ornaments and metrical effects in connexion with the expression of the sentiments. The erotic sentiment demands metaphors and Dīpakas, and prefers the Āryā metre. The heroic affects short syllables, similes, and metaphors; in passages of lively dialogue the metres [702] Jagatī, Atijagatī, and Saṁkṛti; in scenes of battle and violence, the Utkṛti. The sentiment of fury adopts the same metres, and also favours short syllables, similes, and metaphors. The Çakvarī and Atidhṛti metres are appropriate to the pathetic sentiment; it prefers long syllables, a liking shared by the sentiment of horror.

An effort is made by later writers on poetics to apply to the doctrine of the sentiments the theory of excellencies (guṇa), which is laid down generally in Daṇḍin, Vāmana, Bhoja, and other writers. In Daṇḍin [703] we find assigned to the Vaidarbha style a miscellaneous number of qualities, ten in all, which are defined in terms sometimes vague and unsatisfactory; these qualities include both those of sense and sound (artha and çabda). They include strength or majesty (ojas), elevation (udāratva), clearness (prasāda), precision of exposition (arthavyakti), beauty or attractiveness (kānti), sweetness or elegance (mādhurya), metaphorical language (samādhi), and in the use and combination of sounds homogeneity (samatā), softness (sukumāratā), and a natural flow (çleṣa). The chief opponent of the Vaidarbha style is given as the Gauḍa; it is vaguely credited with the possession of features the opposite of those of its rival; more specifically, we find it credited with the fondness for the use of long compounds both in prose and verse, while the Vaidarbha objects to such compounds in verse at least, and with affecting alliterations. Vāmana [704] develops the doctrine by distinguishing ten qualities of sense and ten of sound, and he ascribes all the qualities to the Vaidarbha style; to the Gauḍa he allots reliance on force and beauty, to the exclusion of sweetness and softness, while he recognizes as a third style the Pāñcāla, which is marked with sweetness and softness, and therefore is rather feeble. In Mammaṭa [705] and later we find a new view of the qualities; those of sense are explained away as being rather the absence of defects (doṣa), so that the qualities are reduced to the sphere of sound alone. In this regard they are further reduced from ten to three, sweetness, strength, and clearness, and these are now brought into effective connexion with the sentiments.

Sweetness, the source of pleasure, causing as it were the melting of the heart, is appropriate in the sentiments of love in enjoyment, pathos, love in separation, and calm; it is normal in love in union, and rises in degree successively in the other three forms of sentiment; unmixed in the others, in that of calm it is combined with a small degree of strength, because of the relation of the sentiment of calm to the emotion of disgust. Strength causes the expansion of the heart; it rises in vehemence in the sentiments of heroism, horror, and fury, and it is found also in that of terror. The quality of clearness is appropriate to all the sentiments, and is that which causes the sense to become intelligible, pervading the mind as fire does wood or water a cloth, as the outcome of merely hearing the words. The precise mode in which sweetness is produced is by the use of mutes other than cerebrals, with their appropriate nasals, r and ṇ with short vowels, and no compounds or short compounds; strength results from the use of compound letters, doubled letters, conjunct consonants of which r forms part, cerebrals other than ṇ, palatal and cerebral sibilants, and long compounds. The older names, Vaidarbha, Gauḍa, and Pāñcāla are now given up in favour of refined (upanāgarikā), harsh (paruṣā), and soft (komalā). But Mammaṭa reminds us that in drama very long compounds are undesirable, a rule ignored largely by the later dramatists.

More important than these technical details, which are illustrated often enough in the verses composed by the later dramatists, and no doubt possess considerable antiquity, is the changed view [706] which brings the qualities in the new sense into relation with the sentiment. Sentiment is the very soul of poetry, and the relation of the qualities to it may be most effectually compared with that of such virtues as heroism to the soul of man. They serve to heighten the effect of the sentiment, and therefore they cannot be considered save in close relation to that sentiment. However soft and sweet the verbal form of a work, none the less it cannot be said to possess the quality of sweetness, unless it has a sentiment to which sweetness is appropriate. To give it the name of sweet, if the sentiment is incompatible with sweetness, is compared with regarding a tall man as brave on the strength of his appearance only. The sounds, therefore, produce the qualities only as instruments, for the real cause is the sentiment, even as the soul is the true cause of the heroism and other qualities of a man.

The case of figures, whether of sound or sense, is somewhat similarly handled; the figures are compared with the ornaments which, placed on a man’s body, and through this union with him, gratify the soul; the figures adorn words and meanings which are parts of poetry by their union with them, and thus serve to heighten the sentiment, provided one exists. If there is no sentiment, through the defective ability of the poet, then the figures serve merely to lend variety to the composition, and even when sentiment exists the figures may fail to be appropriate to it. Both figures and qualities thus are in a very intimate relation with the sentiment, but that does not mean that the two are identical.

From this doctrine, which makes sentiment essentially the main element in poetry, the view of Vāmana, [707] who laid down that style was the soul of poetry and that the qualities give the essential beauty or distinction (çobhā) to a poem, while the figures increase such distinction, is necessarily regarded as inadequate. If the doctrine is interpreted to mean that it is the possession of all the qualities which makes a poem, then all compositions in the Gauḍa and Pāñcāla styles would be denied the rank of poetry; if the presence of a single quality gave the right to the style of poem, then a perfectly prosaic verse passage containing the quality of strength would have to be dubbed a poem, while a stanza containing elegant figures, but no qualities, would be denied that style, which in point of fact is regularly accorded by usage and must be recognized as valid.

As regards language we have, as often in the theory, no explanation of a principle which is laid down as accepted, the divergent use of Sanskrit and Prākrit in the same play. Yet it cannot be held that, when the theory was developed in such works as the Daçarūpa, and very possibly in the Nāṭyaçāstra itself, the usage of the plays could be put down simply to the copying of the actual practice in real life. That such was its origin we may believe in the general way; the Vidūṣaka in the Mṛcchakaṭikā derides a woman using Sanskrit as resembling a young cow with a rope through her nose; but there is evidence that already in the time of the Kāmaçāstra [708] the use of Prākrit was artificial. We are there told that the cultured man about town (nāgaraka) in social meetings (goṣṭhī), should neither confine himself to Sanskrit nor to the vernacular (deçabhāṣā) if he is to win repute for good manners. We have here a sign that matters were already, at the time of the Kāmaçāstra, much in the same condition as in modern India, where the use of Sanskrit terms with the vernacular is a regular sign of education. Now Vātsyāyana tells us clearly that those who frequent such gatherings are hetaerae, Viṭas, Vidūṣakas, and Pīṭhamardas, in short the wits of the court, and to them in the theory is assigned Çaurasenī and kindred Prākrit dialects. We are justified, therefore, in assuming that at Vātsyāyana’s epoch in actual life, as opposed to the conventional existence of the stage, Prākrits were definitely out of employment. The same text includes in the requisites of the knowledge of a hetaera the knowledge of the local speech, and, as there is no doubt of the knowledge of the Andhras as kings by Vātsyāyana, it is interesting to note that in the famous passage in which Somadeva tells of the reason why the Bṛhatkathā [709] was written in Paiçācī he treats as the three forms of human speech contemporaneous with Sātavāhana, whose name shows his connexion with the Andhras, Sanskrit, Prākrit, and the vernacular.

The date of Vātsyāyana thus becomes of interest, but unluckily it is still undefined with any precision. [710] It certainly seems, however, that Kālidāsa was familiar with a text very similar to and perhaps identical with the Kāmaçāstra, and this reasonably may be regarded as giving A.D. 400 as the lower limit of date. That the Kauṭilīya Arthaçāstra has been used by Vātsyāyana gives no precise result, in view of the difficulty of dating exactly that text. But the mention by Vātsyāyana of the Ābhīras [711] and Andhras certainly suggests, taken into conjunction with his silence as to the Guptas, that he wrote before the power of the latter had established itself in western India, and we may assign his work to approximately A.D. 300. If so we must believe that already in Kālidāsa’s age the Prākrits of his characters were more or less artificial, and with this well accords his introduction of Māhārāṣṭrī for the verses of those to whom Çaurasenī is assigned in prose, an obviously literary device.

Elaborate rules for the use of language [712] by the characters are given in the Nāṭyaçāstra and, in much less detail, by the Daçarūpa. The use of Sanskrit is proper in the case of kings, Brahmins, generals, ministers, and learned persons generally; the chief queen is assigned it, and so also ministers’ daughters, but this rule is not in practice observed. On the other hand, it is used by Buddhist nuns, hetaerae, artistes, and others on occasion. It is a rule that in the description of battles, peace negotiations, and omens Sanskrit shall be resorted to, and this is done by Bṛhannalā in Bhāsa’s Pañcarātra. The use of Sanskrit by allegorical female types is also found both early and late.

The general rule for women and persons of inferior rank [713] is the use of Prākrit, but it may be resorted to as a means of self-aid by persons of higher position. The types of Prākrit to be used are described with much confusion in the Nāṭyaçāstra, and the amount of variation contemplated is large. Thus the use of Çaurasenī is permitted in the Çāstra in lieu of the dialect of the Barbara, Andhra, Kirāta, and Draviḍa, though these may be used. The Çāstra gives seven different Prākrits as in use. Çaurasenī is the speech of the land between the Yamunā and the Gan̄gā or Doab; it is to be used by the ladies of the play, their friends and servants, generally by ladies of good family and many men of the middle class. Prācyā is assigned to the Vidūṣaka, but in fact he speaks practically Çaurasenī, and therefore the term can only denote an eastern Çaurasenī dialect. Āvantī is ascribed to gamblers or rogues (dhūrta), but is only an aspect of Çaurasenī, as spoken at Ujjayinī, and the Prākrit grammarian Mārkaṇḍeya calls it a transition between Çaurasenī and Māhārāṣṭrī. Māhārāṣṭrī is unknown to the Çāstra; it is assigned to the verses of persons who use Çaurasenī by the Daçarūpa, while the Sāhityadarpaṇa limits it to the verses of women; normally, but not absolutely, it is used in all verses, [714] though Çaurasenī verses occasionally occur, and possibly were more frequent originally. The earlier drama of Açvaghoṣa and Bhāsa has no clear evidence of Māhārāṣṭrī at all. Ardha-Māgadhī is prescribed for slaves (ceṭa), Rājaputras and guildsmen (çreṣṭhin) by the Çāstra, but, save in Açvaghoṣa and possibly the Karṇabhāra of Bhāsa, it is unknown to our dramas. Māgadhī, on the other hand, is important in theory, and of some consequence in practice; it is ascribed to all those men who live in the women’s apartments, diggers of underground passages, keepers of drink shops, watchers, and is used in time of danger by the hero, and also by the Çakāra, according to the Çāstra. The Daçarūpa assigns it and Paiçācī to the lowest classes, which accords with facts as regards Māgadhī, but Paiçācī is not found clearly in the dramas.

The Nāṭyaçāstra provides for the use of Dākṣiṇātyā in the case of soldiers (yodha), police officers (nāgaraka), and gamblers (dīvyant), and there are slight traces in the Mṛcchakaṭikā of the existence of this dialect. Bālhīkā is assigned by the same authority to the Khasas and the northerners, but has not yet been traced in any drama.

We learn also from the Çāstra and from Mārkaṇḍeya in special of a number of Vibhāṣās, [715] which seem to be modified forms of the more normal Prākrits, as stereotyped for use by certain characters in the drama. Thus the Çāstra attributes Çākārī to the Çakas, Çabaras, and others, while the Sāhityadarpaṇa accords Çābarī to these persons. The Çāstra ascribes Çābarī to charcoal-burners, hunters, wood-workers, and partly also to forest dwellers in general, and Ābhīrī is ascribed with the option of Çābarī to herdsmen, Cāṇḍālī to Caṇḍālas, and Drāviḍī to Draviḍas, while Oḍri, mentioned by the Çāstra, is left unascribed; presumably it was assigned to men of Orissa. Something of this is seen in the Mṛcchakaṭikā, where Çākārī, Cāṇḍālī, and a further speech Ḍhakkī or Ṭākkī appear. They all have nothing very marked as to their characteristics; the first two may be allied to Māgadhī, the last is more dubious.

The addition of Chāyās or translations in Sanskrit to explain the Prākrit is normal in the manuscripts of the dramas, and it is certain that it is old, for it is alluded to by Rājaçekhara in his Bālarāmāyaṇa. Evidently, even so early as A.D. 900, there was no public which cared for Prākrit without a Sanskrit explanation.

On the subject of the use of stanzas, as opposed to prose, the text-books are curiously and unexpectedly silent. [716] This indicates how entirely empirical they are in these matters. The use of Prākrits in the dramas obviously varied, and something had to be said regarding this point, but the alternation of prose and verse is accepted as something established, on which comment is unnecessary. The fact is recognized, but its implications and purpose remain unexplored. In the stanzas themselves, it is clear, we must distinguish between those which were sung and those which were simply recited; recitation must clearly have been the normal form of use, and as sung we have normally at any rate only some of the stanzas in Māhārāṣṭrī which are placed in the mouths of women. Çaurasenī stanzas, on the other hand, we may assume to have been recited, but the distinction has practically vanished from the texts preserved.

7. THE DANCE, SONG, AND MUSIC

Of the part played by the song, dance, and music in the drama the theorists curiously enough tell us comparatively little of interest, though it is certain that both were most important elements in the production of sentiment. The types of dance recognized in the Nāṭyaçāstra are two, the violent dance of men, invented by Çiva himself, the Tāṇḍava, and the tender and voluptuous dance of Pārvatī, the Lāsya. The latter alone, by reason of its special importance, is carefully analysed into ten parts by the Çāstra, [717] which shows the essential union of song and dance. The first is the song proper, which is sung by one seated, to the accompaniment of a lute, without dancing; the recitation standing (sthitapāṭhya) is a declamation in Prākrit by a woman pacing rapidly under the influence of love, or it may also mean, according to Abhinavagupta, a declamation by a woman in anger. The recitation sitting (āsīna) is performed by a woman lying down, under the stress of sorrow, without musical accompaniment. In the Puṣpagaṇḍikā various metres are used; Sanskrit may be employed; men act as women and vice versa, and there is a musical accompaniment. In the Pracchedaka a woman sings to the lute her grief at her lover’s infidelity. The Trigūḍha is the acting of a man in woman’s dress, as of Makaranda in the Mālatīmādhava, Act VI. The Saindhava is a song to a clear accompaniment of a lady whose love has failed to keep his tryst. The Dvigūḍhaka is a harmonious song, full of sentiment, in dialogue form. The Uttamottaka is a song filled with the bitterness of a troubled love. The Uktapratyukta is a duet, in which one lover addresses to the other feigned reproaches. These divisions, of course, appear to ignore their nature as parts of a dance, but it must be remembered that the motions of the performers are essential in the performance.

The music of the drama is not described at length in the later theorists; what is clear is that each sentiment has its special appropriate music, and each action its special accompaniment. Thus the Dvipadikās accompanied the performance of the rôles of persons distressed, unwell, and unhappy; the Dhruvās were chosen so as to intimate at once to the audience the quality of the new arrival on the stage. [718]

8. THE PRELIMINARIES AND THE PROLOGUE

The Nāṭyaçāstra [719] prescribes an elaborate series of preliminaries (pūrvaran̄ga) which must be performed before the actual drama begins; they are intended to secure divine favour for the performance, each act having a definite share in the result, and doubtless they present us with a reminiscence of the early theatre in the mingling of music, dance, and song. First there is the beat of drum (pratyāhāra) announcing the beginning of the performance, and the carpet is spread out for the orchestra; the singers and the musicians then enter and take their places (avataraṇa): then the chorus try their voices (ārambha); the musicians try their instruments (āçrāvaṇā); they tune up their wind and string instruments, and manipulate their hands to make them ready for the work; then an instrumental concert follows, succeeded by the appearance and practice steps of the dancers. [720] A song follows, to please the gods; then the Tāṇḍava is performed, increasing in violence as it proceeds; then a song accompanies the raising by the Sūtradhāra of the banner (jarjara) of Indra; he scatters flowers and purifies himself with water from a pitcher borne by an attendant, while another carries the banner; there follows a perambulation of the stage, the praise of the world guardians, and homage to the banner. Then comes the Nāndī or benediction; it is followed by the recitation by the Sūtradhāra of a verse in honour of the god whose festival it is, or the king or a Brahmin; then comes the Ran̄gadvāra, which is said to mark the beginning of the dramatic action (abhinaya), the Sūtradhāra reciting another verse, and bowing before the banner of Indra. There follow steps and movements of erotic character (cārī) in honour of Umā, and more violent movements of the same kind in honour of the Bhūtas. A discussion (trigata) between the Sūtradhāra, the Vidūṣaka, who talks nonsense, and an attendant follows. Finally the Prarocanā announces the content of the drama, and the Sūtradhāra and his two attendants leave the stage, and the preliminaries are ended.

Immediately after, according to the Nāṭyaçāstra, another person, similar in appearance and qualities to the Sūtradhāra, is to enter and introduce the play, a function which gives him the style of introducer, Sthāpaka. [721] His costume should indicate the nature of the drama, as dealing with divine or human affairs. An appropriate song greets his entrance, he dances a Cārī, praises the gods and Brahmins, propitiates the audience by verses alluding to the subject of the play, mentions the name of the author and the play, and describes some season in the verbal manner, thereby opening the prologue (prastāvanā, āmukha, sthāpanā) [722] of the play. The essential feature of the prologue is an address by the director with an attendant (pāripārçvika) or an actress or the Vidūṣaka on some personal business which indirectly hints at the drama. The mode of connexion is given by Dhanaṁjaya as threefold, as in the Nāṭyaçāstra; the words of the director may be caught up (kathodghāta) by a character in the drama, entering from behind the curtain, as in the Ratnāvalī Yaugandharāyaṇa catches up the consolation offered to the actress which is applicable to his own scheme, and in the Veṇīsaṁhāra Bhīma brusquely denounces the benediction of his adversaries. Or a person may enter (pravṛttaka), who has just been mentioned by the director in a comparison with the season of the year, as in the Priyadarçikā. Excess of representation (prayogātiçaya) is taken in the Daçarūpa as applying to a case where the director actually mentions the entry of a character of the drama, as at the beginning of the Çakuntalā, where he assures the actress that her song has enchanted him, as the gazelle enchants Duḥṣanta, who just then enters. Viçvanātha, on the other hand, treats this form as an instance of continuance (avalagita), and interprets the phrase as denoting the supersession of the director’s action; thus, in the lost Kundamālā, about to call on the actress to dance, he hears the word, ‘Lady, descend’, and realizes that it is a reference to Sītā, who is being led into exile. He admits also the abrupt dialogue (udghātya) as a means of connexion; thus in the Mudrārākṣasa the director alludes to the demon of eclipse as eager to triumph over Candra, the moon, and Cāṇakya behind the scenes calls out, ‘Who then while I live claims to triumph over Candragupta?’ and enters a moment later. The theorist Nakhakuṭṭa is also credited with the view that a voice behind the scenes or from the air may be used to introduce the chief personage.

This account of the preliminaries and the prelude presents obvious difficulties both in itself and in connexion with the actual specimens of the Sanskrit drama. The Daçarūpa and Viçvanātha alike give no details of the preliminaries, and the Nāṭyaçāstra indicates that, in addition to the complete form of Pūrvaran̄ga, there might be an abbreviated form and also an extended form with additional ceremonials. There is an obvious overlapping between the Pūrvaran̄ga and the rest of the performance, for the last element of the former, the giving the content of the drama in the Prarocanā, is essentially an element in the latter. We are quite definitely told by Viçvanātha that in his time there was not a complete performance of the preliminaries; when, therefore, we find in Bhāsa’s dramas that there is no mention of the name of the author or the drama in the prologue, we may safely assume that it was after his time that the practice grew up of transferring from the preliminaries, which were not a matter for the poet, the substance of the Prarocanā, and embodying it in the poet’s own work. In Viçvanātha’s time also we are told that the Sūtradhāra or director performed the whole of the work assigned in the theory to him and the Sthāpaka. But it is extremely difficult to say how far back this goes; the extant dramas with occasional exceptions, [723] such as Rājaçekhara’s Karpūramañjarī and Mādhava’s Subhadrāharaṇa mention only the Sūtradhāra, and Pischel [724] suggested that it was Bhāsa who banished the Sthāpaka, in view of the reference in Bāṇa to his dramas as begun by the Sūtradhāra. It is uncertain, however, what precisely the sense of this reference is. The Daçarūpa expressly provides for the activity of the Sthāpaka, but then proceeds to style him Sūtradhāra, and there is agreement that he is to have the attributes of the Sūtradhāra, so that the use of the name may merely be explained by this reason. This is certainly supported by the express reference in the Sāhityadarpaṇa to the transfer of his functions to the Sūtradhāra and the silence of the Daçarūpa on this head. The point would be of importance only if it meant that Bhāsa dropped the Pūrvaran̄ga as part of the drama; nothing, however, even hints at this; as we have seen, his omission to name himself or his play in the prologue tells strongly in favour of the view that the old Prarocanā was still in use.

More complex still is the question of the Nāndī or benediction. Most Sanskrit dramas open with a verse or verses of this type, followed by the remark, ‘At the close of the Nāndī the Sūtradhāra enters,’ but in Bhāsa’s dramas, in old manuscripts of the Vikramorvaçī, and now and then in South Indian manuscripts of such plays as the Nāgānanda, the Mudrārākṣasa, and other more modern dramas, [725] we find the play begun with these words, and a verse or verses following. We have also the direct testimony of Viçvanātha, who tells us that some authorities held that the introductory verse in the Vikramorvaçī which normally passes for the Nāndī was not that at all, but was the Ran̄gadvāra, with which, according to the Nāṭyaçāstra, the play properly begins, as in it we first find acting in the shape of a combination of speech and action; that verse, they argued, could not be reconciled with the definition of the extent of the Nāndī given in the Nāṭyaçāstra; others, however, on the authority of Abhinavagupta repelled this objection. Viçvanātha adopts as the definition of Nāndī what is recited in praise of a deity, Brahmin, king or the like, and is accompanied by a benediction, consisting of twelve inflected words (with nominal or verbal endings) or eight lines (quarter-verses); this would exclude the beginning of the Vikramorvaçī, but Abhinavagupta permits of a greater variety of forms. In Viçvanātha’s view the Nāndī is part of the preliminaries, which must be preserved, however much these are shortened. It is clear, therefore, that gradually the benediction, like the Prarocanā with its appeal to the benevolence of the audience, [726] came to be worked into the play by the author himself, though the period when the custom became normal cannot be stated with any precision, and in the south of India, at any rate, the older practice of leaving the benediction to the Sūtradhāra seems to have been sometimes followed. There can, indeed, be little doubt that the extent to which the preliminaries were retained differed from time to time; Viçvanātha evidently contemplates their almost total disappearance, but the Amṛtodaya of Gokulanātha in the sixteenth century assumes their presence; the authority of the Nāṭyaçāstra told heavily in their favour, and the stock phrase, ‘Enough of this ceremony,’ which occurs frequently at the opening of the plays, doubtless refers to the dance, song, and music with which the drama was prefaced. [727]

These facts explain the confusion [728] of the notices of the theorists as to the actor by whom the benediction is to be recited. We find ascribed to Bharata the view that a special actor, the Nāndī, should recite it, or that duty should be performed by the Sūtradhāra; another authority permits the Sūtradhāra or any other actor to recite it. The situation is complicated by the rule that at the end of the preliminaries the Sūtradhāra is supposed to leave the stage and the Sthāpaka to come on, while our dramas, as a rule, have the benediction followed by the entry of the Sūtradhāra, or rarely, as in the Pārthaparākrama, the Sthāpaka. The theory, therefore, suggests that the benediction is recited by the Sūtradhāra or Sthāpaka (called Sūtradhāra by reason of similarity of function and character) behind the curtain, and then he enters on the stage. The matter is not cleared up by the practice followed in the embryo dramas introduced into others: in that included in the Bālarāmāyaṇa the Sūtradhāra recites a benediction of twelve inflected words, and then proceeds with the prologue without a break; in the Jānakīpariṇaya it is one of the actors who does so, as in Ravivarman’s Pradyumnābhyudaya, the director then beginning the play; in the Caitanyacandrodaya the benediction is recited behind the curtain, but that is stated to be because the piece to be acted is a Bhāṇa or Vyāyoga, implying that in other cases it normally was recited on the stage, presumably by an actor other than the director.

The extent of the benediction was, as we have seen, disputed. [729] Bharata’s rule of eight or twelve Padas does not stand alone, for he is credited with mentioning four or sixteen as possible numbers, and Pada may mean inflected word, line, or proposition. Abhinavagupta allows three, six, or twelve Padas in a benediction of three times; four, eight, or sixteen in one of four times; and definitely takes Pada as proposition; illustrations of eight- and twelve-Pada benedictions of this type are given by Abhinavagupta and Bharata. The dramas differ; the Çakuntalā has one of eight propositions or four lines; the Ratnāvalī four stanzas; the Mālatīmādhava and the Mudrārākṣasa eight lines each; the Uttararāmacarita twelve words.

Harmony between the benediction and the character of the drama is naturally demanded by the theory, and is observed largely in practice; thus the Prabodhacandrodaya, a philosophic drama, begins with an adoration of the sole reality, the Mudrārākṣasa, a drama of political intrigue, with a verse as tortuous as the diplomacy of Cāṇakya. It is a characteristic of the determination to carry matters to extremes which distinguishes Indian theory that attempts are made to extract from the benediction not merely a general harmony with the theme, but also a reference both to the main characters and to the chief events. [730]

9. THE TYPES OF DRAMA

The types of drama are distinguished by the theorists according to the use which they make of the various dramatic elements enumerated. The highest of the ten main forms, Rūpakas, is the Nāṭaka or heroic comedy. The term is generic; it may denote any representation whether by pictures or dumb show, but it has also the more important specific sense of the drama proper.

The subject of a Nāṭaka [731] should be drawn from tradition, not invented; the hero should be a king, royal sage, or god, who may appear in human form; the dominant sentiment must be the heroic or the erotic, but all may be illustrated, and that of wonder is well suited for the dénouement, which should be led up to through the whole series of stages of the action and junctures. The end must be happy; tragedy is forbidden, though the prohibition is unexplained. The prose should be simple without elaborate compounds; the verses clear and sweet; the Prākrits should be varied; the whole style noble and harmonious, with full use of all the beauties and the adventitious attractions of the song and the dance as well as music. The number of acts should be from five to ten; if a play contains every kind of episode, it is styled a Mahānāṭaka, if it has ten acts. The rule is generally obeyed, but late dramas styling themselves Nāṭakas are known of one (Ravidāsa’s Mithyājñānaviḍambana), two (Vedāntavāgīça’s Bhojacarita), three, or four acts, [732] and one comparatively early work exists in one version of fourteen acts, without any passage in Prākrit, the Mahānāṭaka; the Adbhutārṇava of a Kavibhūṣaṇa has twelve acts. The name of a Nāṭaka should be derived from the hero or the subject-matter, and this is regularly the case. Four or five is the number of chief personages permitted.

The bourgeois comedy, Prakaraṇa, [733] is a comedy of manners of a rank below royalty, which in the main follows the laws of construction of the Nāṭaka. The subject-matter is to be framed at his good pleasure by the poet. The hero should be a Brahmin, minister, or merchant, who has fallen on evil days and is seeking through difficulties to attain property, love, and the performance of duty, in which he at last succeeds. The heroine may be of three types, a lady of good family, as in the lost Puṣpadūṣita (°bhūṣita); a hetaera as in the lost Taran̄gadatta; or a lady of good family may share the honours with a hetaera, with whom, however, she may not come in contact, as in the Cārudatta and the Mṛcchakaṭikā. The drama offers an appropriate place for slaves, Viṭas, merchant chiefs, and rogues of various kinds. The erotic sentiment should dominate, though Dhanaṁjaya allows also the heroic, and the structure should include all five junctures. The number of acts should be as in the Nāṭaka, and the name be derived from the hero or heroine or both, as in the Mālatīmādhava and the Çāriputraprakaraṇa of Açvaghoṣa. It must, however, be noted that the Pratijñāyaugandharāyaṇa has but four acts, and the Mṛcchakaṭikā, unlike the Cārudatta, does not follow the rule as to name.

The supernatural drama, Samavakāra, [734] is described in our sources obviously on the basis of a single play, the Amṛtamanthana, the churning of the ocean to obtain the ambrosia, at which all participants attained their desires. The precise duration of each of its three acts is given, at twelve, four, and two Nāḍikās (of forty-eight minutes). The subject must be taken from a tale of the gods and demons. The juncture, pause, is omitted, and the expansion (bindu) as an element of the plot. The number of heroes may reach twelve, each pursuing an object which he attains. The heroic sentiment dominates. Each act exhibits one type of cheating, tumultuous action, and love. The graceful manner is excluded, or but faintly developed; the Uṣṇih, Kuṭila, and Anuṣṭubh metres are appropriate. The description fits but loosely Bhāsa’s Pañcarātra, the only old drama to which that name may plausibly be applied.

The Īhāmṛga, [735] of which no old example is known, owes its name, according to the Daçarūpāvaloka to the fact that in it a maiden as hard to attain as a gazelle (mṛga) is sought after (īhā). The subject is one partly derived from legend and partly the poet’s imagination; in special, if the legend relates the death of a great man, this result must be avoided. The essence of the drama is that some one seeks to deprive the hero, who on one view may be divine or human, on another divine only, of a heavenly maiden; the result is a conflict of wills, but actual fighting is to be avoided by artifice. The hero and his rival must both be of the noble and haughty type; the latter must do wrong in error. Only the first two and the last junctures are allowed, and the graceful style is excluded. There are four acts, but Viçvanātha mentions a view which allows one act only and makes the hero a god, or six rivals for a divine maiden’s hand.

The Ḍima [736] is also little known, though the Nāṭyaçāstra cites a Tripuradāha as a specimen. Its subject is to be legendary; there is to be no pause juncture. The heroes are sixteen gods, demi-gods, and demons, all of the haughty type; magic, sorcery, combats, eclipses of the sun and moon are in place. The erotic and comic sentiments are excluded, that of fury is predominant. There are four acts without introductory scenes of any kind, but the late Manmathonmathana of Rāma has them. The graceful manner is forbidden. It is clear that the type is described on the basis of inadequate material; it may represent a popular form of entertainment which did not attain full recognition. The origin of the name is unknown, for no root ḍim, to wound, is found in the language, though Dhanika asserts its existence.

The Vyāyoga [737] is, as its name suggests, a military spectacle. Its subject must be legendary, its hero a god or royal sage, but Dhanaṁjaya allows a man. It is in one act, the action not extending over a day, and it is filled with strife and battle, the intervention of women as the cause of battle being excluded. The first two and last junctures alone are permitted, the erotic and comic sentiments are barred, and the graceful manner. The type is old, for it is found in Bhāsa and revives later.

The Act or Isolated Act (An̄ka, Utsṛṣṭikān̄ka) [738] is a single-act piece, whose longer style serves to discriminate it from an act of a normal drama. Its subject is taken from legend, but may be developed by the poet; the first and last junctures alone are permitted. The hero should be human, of the common folk, according to the later theory. The sentiment should be the pathetic, and the style the verbal. The laments of women should accompany the description of battles and fights, but these should not take place on the stage. Viçvanātha gives the Çarmiṣṭhāyayāti as an example, but the type is not represented by any early play.

The farce, Prahasana, [739] on the other hand, has every sign of popular origin and vogue. The subject is the poet’s invention; it deals essentially with the tricks and quarrels of low characters of every kind. There is but one act, and only the first and last junctures; the comic sentiment predominates. The Daçarūpa recognizes three kinds; the pure is that in which heretics, Brahmins, men- and maid-servants and parasites are represented in appropriate costume and language; the modified represents eunuchs, chamberlains, and ascetics in the garb, and with the speech, of lovers; and the mixed is styled so because it contains the elements of the Vīthī, and is filled with rogues. Only the first and last are recognized by the Nāṭyaçāstra, the second being included in the third, while Viçvanātha recognizes the possibility of there being only one hero or several, and allows the use of two acts in such a case, as in the Laṭakamelaka. The graceful and violent manners are excluded.

The monologue, Bhāṇa, [740] has also an obviously popular character and origin. The subject-matter is invented by the poet; a parasite sets forth his own or another’s adventures, appealing to both the heroic and the erotic sentiments by descriptions of heroism and beauty in the verbal manner. There are only the first and last junctures, and but one act. The actor speaks in the air, repeating answers supposed to be received. The elements of the Lāsya are specially in place, a fact which shows that we have here a formal version of a primitive mimetic performance. Viçvanātha gives as example the Līlāmadhukara; the Çāradātilaka is one of the best known.

The garland, Vīthī, [741] has a certain similarity to the Bhāṇa in that it includes frequent speeches in the air, and is in one act. But it is played by one or two actors, or, according to Viçvanātha on one view found in the Nāṭyaçāstra, by three, one of each station in rank. The leading sentiment is the erotic, but others are hinted at. The graceful manner is forbidden by the Nāṭyaçāstra, but enjoined by the other authorities, and the elements of the garland are available. Only the first and last junctures are employed, but all the elements of the plot are present. The theorists are sadly at a loss to explain the name garland; it is suggested that the several sentiments are gathered into it as into a garland, or the meaning ‘way’ or ‘road’ is accepted in lieu. The only example given by Viçvanātha is the Mālavikā, which is not the Mālavikāgnimitra; the first act of the Mālatīmādhava is styled Bakulavīthī, but is in no sense even taken by itself an example of this type.

The later theory as seen in Viçvanātha [742] adds descriptions of eighteen minor forms of drama, Uparūpakas, which represent refinements on the original scheme. Needless to say, though omitted in the Nāṭyaçāstra, quotations are found ascribing to Bharata the doctrine, though he mentions in them but fifteen with several variations of name; [743] the Agni Purāṇa [744] mentions eighteen with some variants of name, while a verse cited by Dhanika [745] names seven forms of mimetic dramas, which it classes in conjunction with the Bhāṇa. The age of these divisions is, therefore, uncertain; the Daçarūpa condescends to mention only the Nāṭikā, but obviously knows of the existence of others, confining its scope to the main forms, as its title indicates.

The Nāṭyaçāstra [746] mentions, in a passage suspected of interpolation, but without special cause, a type of dramas Nāṭī, which later is styled Nāṭikā, or lesser heroic comedy. The subject-matter in this view may be either legendary or invented; the later opinion requires it to be invented as in the Prakaraṇa, which is the model for the Nāṭikā in this regard. The hero is to be that of the Nāṭaka, a gay king, and the intrigue consists of his efforts to attain marriage with the heroine, who is an ingénue of royal family, whom he is destined to marry, but who by some accident or design has been introduced into the harem in an inferior capacity. The lovers have to strive against the jealousy of the queen, a lady of mature character and devotion to the king, who at last is induced to sanction the nuptials. The life of the court gives opportunity for introducing music, song, and the dance as elements in the entertainment. The graceful manner is appropriate, and the erotic sentiment is prescribed; by an excess of zeal, when the drama as usual has four acts, they are in theory to contain each one of the four members of the graceful style. A lesser number of acts is allowed by Dhanaṁjaya. There is certainly not much difference between such a Nāṭaka as the Mālavikāgnimitra and the normal Nāṭikā, save the length, as expressed in the number of the acts, but it would be unwise to assert that the distinction is based on this alone. It is a fact that both in the Priyadarçikā and the Ratnāvalī the poet has freely enough invented his episodes, and this is a fact justifying the discrimination.

The little bourgeois comedy, the Prakaraṇikā, [747] is precisely of the same character as the Nāṭikā, save that its hero and heroine are of the merchant class. It is clear that it is due merely to a false desire for symmetry, as it is merely a Prakaraṇa when judged by the three determinants of plot, character, and sentiment, and Dhanika rightly rejects it as a species, though Viçvanātha admits it.

A variant of the Nāṭikā is the Saṭṭaka, [748] which differs from it merely by being all in Prākrit, in having no introductory scenes of any kind, and in having the acts called Javanikāntara. As the name denotes a form of dance, it is quite possible that it owes its origin as a species to the use of such dances in these plays. We have an example in Rājaçekhara’s Karpūramañjarī.

The Troṭaka [749] or Toṭaka is merely a variant of the Nāṭaka; the Bengālī recension of the Vikramorvaçī which contains Apabhraṅça verses and an appropriate dance of the distracted king alone gives the name. The term denotes both a dance and confused speech, and the origin of the species need be sought only in this peculiarity. The other manuscripts call it a Nāṭaka.

The other species enumerated have no representatives in the old literature, nor is this wonderful, for they show the character rather of pantomime with song, dance, and music than of serious drama; the Goṣṭhī [750] has nine or ten men and five or six women as actors; the Hallīça [751] is clearly a glorified dance; the Nāṭyarāsaka [752] a ballet and pantomime; the Prasthāna, [753] in which hero and heroine are slaves, is based on a mimetic dance; so also apparently are the Bhāṇikā, [754] or little Bhāṇa, and the Kāvya, both one-act pieces; the Rāsaka, of the same general type, includes dialect in its language. The Ullāpya may have one or three acts, and its hero is of high rank, while battles form part of its subject, as they do also in the Saṁlāpaka, which may have one, three, or four acts. The Durmallikā has four acts, a hero of low rank, and a precise time-table of duration of acts. The Vilāsikā has one act, but is interesting in that the hero has, to support him, not only the Vidūṣaka, but also the parasite and a friend (pīṭhamarda); the sentiment is erotic. The Çilpaka is mysterious, for it has four acts, allows all the manners, has a Brahmin as hero with a man of lower rank as secondary hero, excludes the calm and comic sentiments, and has twenty-seven most miscellaneous constituents; if a pantomime, it was clearly not amusing. The Pren̄khaṇa, or Prekṣaṇa, is a piece in one act, with a hero of low birth, full of combats and hard words; it has no introductory scenes, and both the benediction and the Prarocanā are performed behind the scenes, but none of the late works which bear approximately this title conforms to type. The Çrīgadita is in a single act, the story legendary, the hero and heroine of high rank, the manner verbal; the word Çrī is often mentioned, or the goddess is presented seated and singing some verse. The only play known of that name is the Subhadrāharaṇa of Mādhava before A.D. 1600, which is much like an ordinary play, but contains a narrative verse, suggesting connexion with the shadow-drama. It is characteristic that the theory ignores wholly this type.

10. THE INFLUENCE OF THEORY ON PRACTICE

Though we cannot say precisely at what date the Nāṭyaçāstra obtained definite form, we can be assured that by the time of Kālidāsa it was not merely known, but its authority was already accepted as binding on poets. The mere fact that Kālidāsa’s dramas exhibit a marvellous fidelity to the rules of the Çāstra might be explained by the theory that it drew its principles from them rather than vice versa. But in his epics Kālidāsa, in due accord with the duty of a poet to display every form of his erudition, has emphatically shown a far-reaching competence in the terminology of the Çāstra. In the Kumārasambhava [755] Çiva and Pārvatī watched the performance in honour of their nuptials of a Nāṭaka in which the different dramatic manners were combined with the junctures, the modes of the music corresponded with the sentiments, and the Apsarases displayed their grace of form. There are similar references in the Raghuvaṅça. [756] The knowledge of the Çāstra by later writers goes without saying. The author of the Mudrārākṣasa [757] depicts Rākṣasa as comparing political combinations with the work of a dramatist and giving a brief plan of the structure of the drama, and Bhavabhūti [758] and Murāri [759] alike show familiarity with the terminology of the Çāstra as well as with its rules. The most complete proof, however, of the domination of the theory is the absence of any original creations in dramatic form. There must, it is certain, have been a time when the genius of Indian poetry was active in trying and developing the new instrument of drama, but with the appearance of the Nāṭyaçāstra this creative epoch came to all intents and purposes to a close, and the writers of the classical drama accept without question the forms imposed upon them by authority, although that authority rests on no logical or psychological basis, but represents merely generalizations, often hasty, from a limited number of plays.

The Nāṭaka, accordingly, remains the form of drama par excellence, a pre-eminence due to its comparative freedom from narrow restrictions as well as to the submissive spirit of the dramatists. The form serves very different purposes; it accommodates itself not only to the grace and charm of Kālidāsa, but to the unmeasured and irregular genius of Bhavabhūti; it permits of the political drama of Viçākhadatta, as well as the philosophical disquisitions of Kṛṣṇamiçra and the devotional fervour of Kavikarṇapūra’s Caitanyacandrodaya.

The Prakaraṇa is essentially similar to the Nāṭaka save in the social status of the hero and heroine; the distinction between the Mālatīmādhava and a Nāṭaka is far less important than the similarity. The Mṛcchakaṭikā, indeed, departs from type, but that is not surprising now that it is known that it is based on Bhāsa’s Cārudatta, which is not merely the work of a man of unusual talent, but came into being before the rules of the drama had attained the binding force they later achieved. The Nāṭikā, however, which is likewise closely allied to the Nāṭaka, became stereotyped at an early stage, leaving no room for serious innovation; the charms of the song and dance appear to have prevailed, and to have dissuaded efforts at originality of plot. The Vyāyoga is hardly more than an aspect of the Nāṭaka; the spirit of such works as those of Bhāsa in this genre is reflected in many passages of the Mahāvīracarita and the Veṇīsaṁhāra.

The farce and the monologue, of which we have many specimens in the later drama, are confined to representations of the lower and coarser side of life, but curiously enough they fail entirely to achieve what might have seemed the legitimate aim of a vivid portrayal of the lives and manners of contemporary society; tradition has proved too strong for the dramatists whose works deal with types, not individuals. On the other hand, we find practically no living tradition of the construction of dramas of the other five classes of the theory, Ḍima, Samavakāra, Īhāmṛga, Vīthī, and Utsṛṣṭikān̄ka. We may legitimately assume that these were types erected on little foundation of fact, and that, while the theory could restrict enterprise, it could not induce life in forms which had no real vitality of their own. The mere fact that later poets occasionally patronize these forms is sufficient evidence of the strength of the authority of the Çāstra. It is amazing, however, that we find no serious effort to produce pure comedy; the farce and the monologue may hover on the borders of that form; they certainly never attain it.

To the force of the tradition is presumably to be ascribed the absence of any effort at tragedy, though its absence undoubtedly coincides with the mental outlook of the Indian people and their philosophy of life. Bhāsa has indeed been claimed as a tragedian, but with complete disregard for the facts; there is in fact in his dramas disregard of the rule which objects to death on the stage, but the slain are always evil men, whose death is just punishment; the Ūrubhan̄ga may to us be tragic, but that is because we are not adorers of Viṣṇu who regard with relish the fate of the enemy of that god, the evil Duryodhana. The tragic sentiment is nowhere recognized, for the term (raudra), which is unhappily often so rendered, is the sentiment which is based on anger, and has nothing truly tragic in it. The idea is, indeed, entirely wanting in the theory as it is in the practice.

To the developed thought of India, as it existed during the vogue of the drama, there was little possibility of a realization of the elements of which Greek tragedy is composed. The conception of human activity striving with circumstance, endeavouring to assert itself in the teeth of forces superhuman in power and uncontrollable, and meeting with utter ruin, but yet maintaining its honour, which affords the spring of tragedy in Greece, is alien to Indian thought. Fate is nothing outside man; he is subject to no alien influences; he is what he has made himself by acts in past lives; if he suffers evil he has deserved it as just retribution, and to sympathize with him, to feel the pathos of his plight, is really unthinkable. Death, therefore, by violence is merely a just punishment of crime, and it is a more refined taste than that of Bhāsa which bids us banish from the stage the spectacle of what is no more than an execution, a scene as ill-suited to the decorum and good taste of the serious drama [760] as to the rude merriment of the farce or monologue.

11. ARISTOTLE AND THE INDIAN THEORY OF POETICS

It is natural that contemporaneously with the effort to prove the Greek origin of the Indian drama efforts [761] should have been made to establish the indebtedness of the Nāṭyaçāstra to Aristotle’s theory of drama. [762] There is no doubt of the many parallels between the two theories. The unity of action is fully recognized in the Çāstra, and the rule which insists that the events described in an Act shall not exceed in duration a day has a certain similarity to the unity of time in Aristotle, [763] and is much more significant than such agreement as there is as to unity of place. The doctrine that the drama is an imitation (anukṛti) does not differ from the doctrine of Mimesis, but there is an essential distinction in what is imitated or represented; in the Çāstra it is a state or condition, in Aristotle it is action, a distinction absolutely in accord with the different geniuses of the two peoples. The importance of acting is common in both schemes, but Aristotle makes little of the dance. Both stress the plot, which the Çāstra recognizes as the body of the drama. The Indian division of characters as high, middle, and low has a certain parallelism to the Aristotelian distinctions of modes of depicting character as ideal, real, and inferior. The Çāstra, like Aristotle, shows appreciation of the distinction between male and female characters. To some degree we find in the Çāstra the recognition of the necessity of conflict in drama, and of the emotions of pity and fear in the sentiment of pathos and in the element of the development known as Vidrava. The Çāstra also touches on the relation of the feelings aroused in the actor and in the audience as in the Poetics. Both recognize the use of significant names, and deal with the linguistic aspects of style.

Other suggestions of Greek influence may also be adduced; thus we have the mention of what seems a derivative of the Greek caryatides in the description of the theatre; the monologue may be based on the Greek Mime, and we have the actual mention in a passage of the Çāstra of Yavanas, while the description of the Viṭa suggests derivation from the Greek parasite. But it is impossible to take these pieces of evidence as conclusive proof of borrowing; we are, in fact, faced with the usual difficulty that, if there were borrowing, the Indian genius has known how to recast so cleverly and to adapt what it borrowed so effectively that the traces which would definitely establish indebtedness cannot be found. In all the instances enumerated there is no doubt similarity, but there is also essential difference such as renders independent development of the Indian doctrine at least as probable as borrowing.