Chapter VI: Part I: The Origin of the Sanskrit Drama (3)

The actual points of contact between the New Comedy and the Sanskrit drama are, however, scanty. The division of both the Roman drama [123] and the Sanskrit into acts, distinguished by the departure of all the actors from the stage and the number of five as normal, though often exceeded in India, are facts which need not be more than casual coincidences: the divisions in the Sanskrit drama rest on an analysis of the action which is not recorded in Greece or Rome. There is similarity in the scenic conventions, in the asides, in the entry and exit of characters, more notably in the practice that the advent of a new character is usually expressly notified to the audience by a remark from one of the actors already on the stage. But these are all matters which must almost inevitably coincide in theatrical performances produced under approximately similar conditions. Even in the modern theatre with its programmes the necessity of indicating at once the identity of the new comers to the stage is keenly felt.

More value attaches to the argument from the use of Yavanikā, [124] or its Prākrit form Javanikā, for the name of the curtain which covered the tiring room and formed the background of the stage. The word primarily is an adjective meaning Ionian, the Greeks with whom India first came into contact. But it was not confined to what was Greek in the strict sense of the word; it applies to anything connected with the Hellenized Persian Empire, Egypt, Syria, Bactria, and it therefore cannot be rigidly limited to what is Greek. As applied to the curtain it is an adjective, and describes doubtless the material of the curtain (paṭī, apaṭī) as foreign, possibly as Lévi suggests, Persian tapestry brought to India by Greek ships and merchants. The word Yavanikā has no special application to the curtain of the theatre, as would be the case, if it were borrowed as a detail of stage arrangement from Greece. Nor in fact was there any curtain in the case of Greek drama, so far as is known, from which it could be borrowed; Windisch’s contention merely was that the curtain was called Greek because it took the place of the painted scenery at the back of the Greek stage.

As little can any conclusion of Greek borrowing be drawn from the Yavanīs, Greek maidens, who are represented as among the body-guard of the king; [125] for this the Greek drama offers no parallel; it represents the fondness of the princes of India [126] for the fascinating hetaerae of Greece, and the readiness of Greek traders to make the high profits to be derived from shipping these youthful cargoes.

The points of resemblance in regard to the plot are of interest. There is some similarity between the stock theme of the Nāṭikā, the love of a king for a maiden, hindered by various obstacles, and finally successful through events which reveal her as a princess, destined for him in marriage but concealed in this aspect by some accident, and the New Comedy picture of the youth whose affection for a fair lady, apparently of status which forbids marriage by Attic law, but in reality of equal birth, is finally rewarded by the discovery of the mark which leads to her identification. The use of a mark of recognition is undoubtedly common in both dramas. We have in the Çakuntalā the ring [127] which gives part of the title of the play Abhijñāna-Çakuntalā, and in the Vikramorvaçī the stone of reunion (saṁgamamaṇi) which enables Purūravas to recognise his beloved despite her change into a creeper. In the Ratnāvalī we have the necklace which permits the identification of the heroine; in the Nāgānanda, the jewel which, falling from the sky, denotes the fate of the prince; in the Mālatīmādhava the garland plucked by Mādhava, worn by Mālatī, which Saudāminī produces at the dénouement as a sign of recognition; and in the Mṛcchakaṭikā the clay cart in which are placed the jewels used as evidence against the hero. In the same general category fall the ring of the queen in the Mālavikāgnimitra, which the Vidūṣaka obtains from her in order to cure a snake-bite, and employs to bring about the release of Mālavikā; the arrow of Āyus, in the Vikramorvaçī, which reveals to Purūravas his son; and the seal of Rākṣasa in the Mudrārākṣasa of which Cāṇakya makes use to confound his schemes. In some cases the similarity of use of these emblems is close; Mālavikā, taken away by brigands, and Ratnāvalī, rescued from the sea, are real parallels to the heroine of the Rudens, stolen from her father by a brigand, sold to a leno and wrecked on the Sicilian coast, whose recognition is brought about by the discovery of her childish ornaments.

These are striking facts, and the only way to meet them is to show that the motifs in Sanskrit drama have an earlier history in the literature, and can, therefore, be regarded as natural developments. The difficulty presented here is that the literature available consists either of tales, which in any form available to us are later than the period of the supposed Greek influence, or the epic which is of uncertain date, so that no strict proof is available that any of its minor issues antedates the Christian era. But we do find in the epic indications that it was not necessary for Greece to give to India the ideas presented in the drama. The story of the love of Kīcaka for Draupadī, when disguised as handmaiden she served Sudeṣṇā, wife of the king Virāṭa, has a tragic outcome, for his love is repulsed, but it has undoubted affinities with the plot of the Nāṭikā. In the case of the old tale of Nala and Damayantī, the heroine is more happy, for, when separated from her husband who has abandoned her in the distraction of losing his kingdom at dice, she lives in peace, guarded securely from interference; at last she is recognized by a birthmark. In the Rāmāyaṇa the use of signs of this sort is extended to artificial modes: Sītā, stolen away from Rāma, drops her jewels to the ground; the monkeys bear them to their king, who hands them to Rāma, and the hero thus knows beyond a peradventure the identity of the ravisher. To console her in her detention pending his efforts at rescue he sends Hanumant to her, bearing his messages, and gives him his ring to serve to identify him; Sītā sees it and takes heart. We may admit that such incidents are almost inevitable in a primitive society, in which the means of identification were necessarily material, or personal. Nor in the Sanskrit drama is there any preponderant use of this factor; the letter and the portrait [128] are other means, the use of which is recognized in the theory.

The evidence of borrowing based on the Mṛcchakaṭikā by Windisch requires reconsideration in the light of the facts now known regarding the authority of that drama for the early Sanskrit drama. To Windisch it seemed to present every appearance of an early age, and to show close relations to a Greek model. The title he compared with the Cistellaria, ‘little chest’, or the Aulularia, ‘little pot’; the mixture of a political intrigue and a love drama with the mention—only incidental however—of political events contemporaneous with the action in Plautus’s Epidicus and Captivi; the court scene he held to be of Greek inspiration; the meeting of Cārudatta and Vasantasenā he compared with that of the hero and heroine of the Cistellaria; the theft of Çarvilaka, in order to buy the freedom of the slave girl he loves, to the dishonest means adopted by the hero in the new comedy to procure means to purchase his inamorata; the setting free of the slave by Vasantasenā with the attaining of the position of a freedwoman in the Greek drama; finally the elevation of Vasantasenā to the rank of a woman of good character to permit of her legal marriage to Cārudatta is compared with the discovery in the Greek drama of the existence of a free status as the birthright of the maiden whom the hero loves. The Mṛcchakaṭikā, however, is not an early representative of the Indian drama in the sense held by Windisch; it is based on the Cārudatta of Bhāsa, in which there is no mingling of the political and love intrigue, at any rate as we have that play; the title Mṛcchakaṭikā, which departs from the usual model, was probably deliberately chosen to distinguish the new drama from the old. The plays cited have no real combination of political and love intrigues, and the other parallels are far too vague to be taken seriously. The raising of Vasantasenā to a new status is an extraordinary event, which is dependent on an action of the new king Āryaka, who, as an overthrower of the former monarch, exercises the supreme right of sovereignty in favour of the lady, in defiance of the rules of caste. The political intrigue thus becomes a vital element in the play.

Nor can any special value be ascribed to the rule, which is laid down in the theory, and observed in practice, and which confines the events in an act to the limits of a single day, as compared with the rule of Aristotle [129] that the events of a drama should not exceed, or only by a little, the duration of a day. If the rule was borrowed, it was greatly changed in sense by permitting long periods, up to a year, to elapse between the acts in the Sanskrit drama, and the mere moral needs of the approximation to reality requisite for illusion would produce the state of the Sanskrit drama without external influence.

The characters of the drama present problems which are not solved by the theory of borrowing. The figure of the queen, loving her husband, noble and dignified, is compared by Windisch with that of the matrona of the Roman comedy, while her attempts to prevent the union of her husband and the new love are compared to the efforts of the senex to dissuade his son from a rash marriage or intrigue. But it is clear that the comparisons are idle; the rivalry of the old love and the new is an incident of the life of the harem inevitable in polygamy, while it affords an admirable opportunity for the poet to depict the contrast of types and the different aspects of love, his chief theme. Windisch, however, lays most stress on his comparison of the three figures of the Viṭa, Vidūṣaka, and Çakāra, with the parasite, the servus currens, and the miles gloriosus of the Greek drama, and his arguments have a certain weight. It is true that these three, with the Sūtradhāra and his assistant, are given by the Nāṭyaçāstra in a list of actors, and that the five correspond fairly closely with the male personnel of a Greek drama; it is also true that, while Kālidāsa and the Mṛcchakaṭikā with the Cārudatta know the Çakāra, he vanishes from the later drama, and the Viṭa shows comparatively little life, suggesting that the Greek borrowings were gradually felt unsuited to India and died a natural death. But the argument is inadequate to prove borrowing. The Viṭa is, indeed, more closely akin to the parasite than to any other character of the Greek or Roman comedy, but the parasite is lacking in the refinement and culture of his Indian counterpart, who is clearly drawn from life, the witty and accomplished companion who is paid to amuse his patron, but whose dependence does not make him the object of insolence and bad jokes. The Vidūṣaka has, in all likelihood, as has been seen, his origin in the religious drama; his Brahmin caste, and his use of Prākrit can best thus be explained. The alternative views all present far more difficulties; the transformance of the slave into a Brahmin is far too violent a change to be credible, while Lévi’s [130] view which makes him a borrowing from the Prākrit drama, which depicted with truth the type of Brahmin who serves as go-between in love affairs, masking his degraded trade under the cloak of religion, renders it unintelligible why the Brahmins should have consented to maintain him in the Sanskrit drama. Equally unconvincing is Professor Konow’s [131] effort to explain him as a figure of the popular drama, which loved to make fun of the higher classes, especially the Brahmins. There was no conceivable reason why the Brahmins should have kept such a figure in a drama which never appealed to the lower classes, and it is significant that there is no trace of a comic figure of the Kṣatriya class, although the populace doubtless was as willing to make fun of the rulers as of the priests. The similarity between the Çakāra and the miles gloriosus is by no means small, but the argument from borrowing is refuted by the reflexion that such a figure can be explained perfectly easily from the actual life of India in the period of Bhāsa and the Mṛcchakaṭikā, when mercenary soldiers must have been painfully familiar to Indians.

The number of actors is certainly not in accord with the Greek practice; not only has Bhāsa large numbers, but the Çakuntalā has thirty, the Mṛcchakaṭikā twenty-nine, the Vikramorvaçī eighteen, the Mudrārākṣasa twenty-four, and it is only in the later and less inventive Bhavabhūti that we find but thirteen in the Mālatīmādhava and eleven in the Uttararāmacarita.

The prologue in both dramas serves the purpose of announcing the author’s name, the title of the play, and the desire of the dramatist for a sympathetic reception, but the Indian prologue is closely attached to the preliminaries, and has a definite and independent character of its own in the conversation between the Sūtradhāra and his wife, the chief actress, so that borrowing is out of the question. Nor does any importance attach to the fact that Çiva, who is in a special sense the patron of drama, is the nearest Indian representative of Dionysos, or that the time of the festival at which plays were often shown was spring, as in the case of the Great Dionysia at Athens when new plays were usually presented. There is similarity between the Protagonist and the Sūtradhāra, for both undertake the leading parts in the drama, but this and other minor points such as can be adduced are of no value as proofs of historical connexion.

Windisch admitted that in regard to the theatrical buildings there was no possibility of comparison, as the Indian theatre was not permanent, but Bloch [132] has endeavoured to show that the Sītābengā cave theatre has marked affinities to the Greek. The attempt, however, is clearly a failure; the construction of the whole is merely that of a small amphitheatre cut out in the rock for a small audience without any special similarity to the Greek theatre of any period.

More recently the tendency of those who seek to find Greek influence in the making of the Sanskrit drama has turned to the mime as the form of art which exercised influence on India, and the older arguments of Windisch have been given a new shape and in part strengthened in this regard. [133] The mime was performed without masks and buskins, as was the Indian drama. Moreover the mime, at any rate in Roman hands, had a curtain (siparium), which may be compared with the curtain of India. There was also no scene painting in the mime; different dialects were used, and the number of actors was considerable. Further, some of the standing types of the mime may be paralleled in the Indian drama; the zēlotypos has some similarity to the Çakāra, the mōkos to the Vidūṣaka.

Some of the arguments adduced against this theory of Reich’s are admittedly untenable. It is impossible to argue as does Professor Konow that the use of the Mṛcchakaṭikā as a work of early date is a mistake, since the oldest dramas preserved are of quite another type and have no similarity with Greek works. True, the Mṛcchakaṭikā is not as old as it was thought, but the Cārudatta can be substituted in lieu, and there are no dramas older than it, save those of the same author and some fragments of Buddhist drama. Nor have we any very satisfactory evidence of a mime in India at an early date, for a mime means a great deal more than the mere work of a Naṭa. But there are adequate grounds for disregarding the theory. The similarity of types is not at all convincing; the borrowing of the idea of using different dialects from the mime is really absurd, and the large number of actors is equally natural in either case. The argument from the curtain is wholly without probative power; as we have seen, the term Yavanikā refers to material only; it would be very remarkable that the term Greek should be confined to the curtain alone, if the stage were really a Greek borrowing, and, last not least, we have no proof that the Greek mime had the curtain. The new form of the theory must, therefore, claim no more credence than the old. We cannot assuredly deny [134] the possibility of Greek influence, in the sense that Weber admitted the probability; the drama, or the mime, may, as played at Greek courts, have aided in the development of a true drama, but the evidence leaves only a negative answer to the search for positive signs of influence.

There are, undoubtedly, certain considerations which a priori tell against borrowing; to judge from the Roman borrowings from Greece and those of France from the classics, the trace of imitation if it were real would be clear and emphatic. But we can hardly place very great faith in arguments from analogy; India has a strange genius for converting what it borrows and assimilating it, as it did in the case of the image of the Buddha which it fabricated from Greek models. More important is the possibility of tracing the sources of the dramas in the epic and the tales, though here the difficulty of dates prevents the demonstration being complete. The epic and undramatic character of the Sanskrit drama is true enough, but not universally applicable, and the argument is liable to be turned by adopting the view that only Greek influence is contended for, not the exclusion of Indian native influences. The typical nature of the characters, adduced by Professor Konow as a point of difference, seems to indicate a forgetfulness that the Greek drama, and especially the New Comedy, is rich in types, and that the mime depicts types. Nor in that comedy do we find any particularly effective heightening of interest or development of the situation from the characters of the persons, or solutions produced without recourse to cutting the knot by artificial means. In all these matters indeed the Indian drama rather is akin to the Greek than otherwise.

6. THE ÇAKAS AND THE SANSKRIT DRAMA

Professor Lévi, [135] whose opposition to Windisch regarding the possibility of Greek influence on the Indian drama has been noted, is himself responsible for the suggestion that the rise of the Sanskrit drama, as opposed to the more popular religious drama in Prākrit, is to be attributed to the Çakas, whose advent to India was one of the causes of the rapid decadence of the Greek principalities in the north-west. The theory is based on a general view of the elevation of Sanskrit to the rank of the language of literature, as opposed to its restriction to use as the learned and sacred language of the Brahmins. The inscriptions, on the whole, show that Sanskrit as an epigraphic language was introduced by Rudradāman whose Girnār inscription of A.D. 150 is wholly in Sanskrit, though Sanskrit appears in part in Uṣavadāta’s inscription of A.D. 124. The Western Kṣatrapas, of Çaka origin, were, he holds, the first to bring Sanskrit down to earth, while not vulgarizing it, as contrasted with the Hindu and orthodox Çātakarṇis of the Deccan who retained Prākrit in their inscriptions down to the third century A.D. The character of the Çakāra may be regarded in this light; in its hostility to the Çakas it reveals a period when either a prince was opposed to the Çaka rule, or the Çaka dominion had just fallen and was fresh in the minds of the people. The Mṛcchakaṭikā may retain a confused version of the events of the second century A.D. A specific connexion between the Çakas and the creation of drama may be seen in the terminology of the Nāṭyaçāstra, and that of their inscriptions. Rudradāman refers to his grandfather Caṣṭana as Svāmin and Sugṛhītanāman, and Svāmin is freely used in the epigraphic records of the kings of the line from Nahapāna (A.D. 78) onwards. Further Rudrasena in A.D. 205, in referring to his royal ancestors, Caṣṭana, Jayadāman, Rudradāman, and Rudrasena, gives them the epithet of Bhadramukha, ‘of gracious countenance’. These terms, Lévi argues, correspond with the use laid down in the Nāṭyaçāstra, which must have borrowed from contemporary official usage. Further, Rudradāman uses the term Rāṣṭriya as applying to Puṣyagupta, who under Candragupta, the Maurya, some four and a half centuries earlier established the reservoir which he had repaired, and this term occurs in the Çakuntalā and the Mṛcchakaṭikā in the sense of brother-in-law of the king, the sense given to it in the Amarakoça, the earliest Sanskrit lexicon of established authority. To these considerations may be added that Ujjayinī, the capital of the Western Kṣatrapas of Mālava, is a centre, round which as a fan radiate the three great literary Prākrits of the drama, Çaurasenī, Māgadhī, and Māhārāṣṭrī, thus accounting for their use, which else would be difficult to explain.

Lévi’s suggestion, which was accompanied by an admission that the Mṛcchakaṭikā or its source was older than he had formerly argued, and that the possibility of Greek influence was thus increased, has been accepted by Professor Konow [136] with the important modification that in face of the fact that the oldest dramas known to us, the fragments of Açvaghoṣa and those of Bhāsa, ignore Māhārāṣṭrī and that Çaurasenī is the normal prose tongue, he accepts Mathurā as the home of the drama, and ascribes it to about the middle of the first century A.D. This view he supports by the fact that the rulers of Mathurā were also Çaka Kṣatrapas, or Satraps, whose control extends back at least to the beginning of the first century A.D.

It may be feared that neither theory will stand critical investigation, however tempting it may be to obtain an exact date for the Sanskrit drama. The discovery of Açvaghoṣa’s fragments shows that the drama has already attained a very definite and complete form, and we really cannot with any probability assume that the creation of drama preceded this by no more than a century. Even a century, however, brings us further back than the middle of the first century A.D., for Konow’s date of Kaniṣka, about A.D. 150, [137] is probably considerably too late, and should be placed fifty years earlier at least. We are thus separated from Rudradāman by a period of 150 years, probably more, and the theory that the Western Kṣatrapas introduced Sanskrit into the drama falls hopelessly to the ground on chronological considerations alone.

The argument from the use of technical terms is clearly untenable. That Rāṣṭriya in Rudradāman’s inscription has the sense of ‘brother-in-law’ is not supported by the slightest evidence, and is most improbable; the term doubtless denotes governor, and the restricted use is a later development. The use of Svāmin as the mode of addressing the king is not recorded in the Nāṭyaçāstra, and to argue that it, being given in the Daçarūpa and the Sāhityadarpaṇa, must be borrowed from Bharata, as Konow does, is quite impossible. On the contrary, Bharata [138] gives the style to the Yuvarāja, or Crown Prince, presumably as distinct from the king. In the extant dramas after Bhāsa it is not used of the king or Crown Prince. Sugṛhītanāman, denoting perhaps ‘whose name is uttered with respect’, has no parallel in Bharata; only in the later theory do we find Sugṛhītābhidha, which, however, is prescribed merely for the address of a pupil, child, or younger brother to a teacher, father, or elder brother, and therefore stands in no conceivable relation to the term used by Rudradāman. Bhadramukha is the address to a royal prince in Bharata; it is used of kings by Rudrasena, and the literature ignores the specific or royal use. The lack of accord is complete and convincing; if the drama had originated under the Western Kṣatrapas of Ujjayinī, it would not have been so flagrantly out of harmony with the official language.

The whole error of these arguments rests in the belief that the drama developed as a Prākrit drama before it was turned into Sanskrit. The same theory has been applied to every department of secular Sanskrit literature without either plausibility or success; the Mahābhāṣya knows Sanskrit Kāvya before any Prākrit Kāvya is recorded. [139] But, apart from this, it is essential to remember that the drama was religious in origin and essentially connected with epic recitations, and that for both reasons Sanskrit claimed in it a rightful place from the inception. It is certain that the recitations known by Patañjali were in Sanskrit, and it is difficult in the extreme to understand how in the view of Lévi and Konow a Prākrit drama proper ever came into being. Before the coalescence of the epic recitation and the primitive mime believed in by Konow, there cannot have been any drama on his own theory; when they coalesced, Sanskrit must have from the first been present.

The discovery of Açvaghoṣa’s fragments undoubtedly helps greatly to bring the creation of the drama very close up to the time of Patañjali, if not to that date. The first century B.C. can with fair certainty be assumed to be the very latest period at which the appearance of a genuine Sanskrit drama can be placed. If indeed Professor Lüders’s former date for Kaniṣka were correct and he were the founder of the Vikrama era of 57 B.C., [140] then the Sanskrit drama must be dated a century at least earlier, and we would have the paradoxical position that on Professor Lüders’s date of Açvaghoṣa he must place the drama at not later than Patañjali, while when dealing with the Mahābhāṣya evidence he doubts the existence of the drama. Professor Lüders has overlooked this dilemma, which, however, we may evade on his behalf by recognising that he erred in assigning to Kaniṣka a date which the evidence available in 1911 already showed to be quite untenable.

7. THE EVIDENCE OF THE PRĀKRITS

The discovery of Açvaghoṣa’s fragments not only disposes effectively of Professor Lévi’s dating of the rise of Sanskrit drama, since he probably preceded Rudradāman by at least half a century, but it casts a vivid light on the question of the Prākrits and Sanskrit. It must be remembered that Açvaghoṣa was the exponent of a faith which had originally insisted on the use of the vernacular as opposed to Sanskrit, and that it is absurd to imagine that it would have occurred to him to use Sanskrit in dramas of Buddhist inspiration and aim, had not the use of that language been established in the drama of the day. This leads us back once more to the conclusion that the drama from the outset was written in part at least in Sanskrit, and that, therefore, it stands in genetic relation with the dramatic recitations described by Patañjali which were in Sanskrit.

That the drama was also in part in Prākrit from the outset seems extremely probable. The mere recitation of the epic indeed did not demand any intervention of Prākrit, but that such recitations by themselves would produce a true drama is most improbable, and we may legitimately hold that it was only the union of these recitations with action from the religious contest that produced the drama. In that contest we may assume that the lower classes were represented and spoke their own language; in the Vedic Mahāvrata we cannot suppose that the Çūdra who contested the right of the Vaiçya to the symbol of the sun spoke in Sanskrit, nor that the Brahmin and the hetaera exchanged their ritual abuse in the classical tongue, or its Vedic antecedent. The religious festival in which Kṛṣṇa appeared as slaying Kaṅsa must similarly have demanded the use of the vernacular by the humbler members of those who took part in it. The fact that Prākrit appears mainly in the dialogue, Sanskrit pre-eminently in verses, strengthens the view that the new drama derived its verse in the main from the epic recitation, its prose dialogue from the religious contest. The two elements never entirely merged; the Vidūṣaka who comes from one side of the religious ceremonial, that which in Greece lies at the basis of comedy as opposed to tragedy, is not a figure normal in the dramas of mainly epic inspiration; but this is not enough to prove that the drama ever in its early days was merely in Sanskrit. It may indeed have been the case; Bhāsa’s Dūtavākya has no Prākrit, and so far the probability is rather for than against it, as an alternative form.

The question how many Prākrits were used in the primitive Sanskrit drama presents difficulties. The obvious conclusion is that the vernacular employed would be that of the region where the drama came into being, and that this was the Çūrasena country is not to be denied. Çaurasenī in fact appears throughout as the normal prose of the drama; it is the language of the Vidūṣaka and the hetaera and normally of all the characters of a play who are born in Āryāvarta, and no other dialect even in theory vies with it in importance. The theory and the practice after Bhāsa ascribe to Māhārāṣṭrī the honour of the language of verses sung by maidens who would in prose speak Çaurasenī. There can be no doubt that this is not primitive, but is a reflex of the growth and development of the fame of the artificial lyric poetry of which we have an anthology under the name of Hāla, perhaps to be ascribed to the third or fifth century A.D. [141]

To what extent any other Prākrit was used in the earliest drama we cannot effectively determine. Bhāsa has only, besides Çaurasenī, Māgadhī of two kinds, and a few hints of what may be styled Ardha-Māgadhī, while Açvaghoṣa has three dialects which suggest much older forms of Çaurasenī, Māgadhī, and Ardha-Māgadhī. The use of these dialects for characters by Açvaghoṣa explains itself naturally from his familiarity with the Buddhist scriptures whose original was very probably in something approximating to the Ardha-Māgadhī [142] he knew, and the fact that the speaker of Old Māgadhī is the Duṣṭa, or bad man, reminds us of the bad character enjoyed [143] by the Magadha. Lévi’s [144] suggestion that the Māgadhī of the drama comes from its epic element, and that the Māgadhas were the reciters of Prākrit epic compositions, is clearly untenable, and indeed seems to have been later abandoned by its author in favour of the suggestion that the Prākrits of the drama were evolved, because the drama was produced at Ujjayinī, which was a meeting place of different dialectical forms. This theory might be revised to adapt it to making Mathurā the headquarters of the drama and Māgadhī and Ardha-Māgadhī the other dialects, but the restricted use of anything but Çaurasenī by Bhāsa suggests that the introduction of other Prākrits was a gradual process. In point of fact it never attained great vitality, and in the developed drama Çaurasenī and Māhārāṣṭrī alone play any real part. The ground for the more extended use of dialects when found may be attributed to literary purposes rather than to any attempt to imitate the speech of the day, as Sir George Grierson [145] has suggested. The ground for this conclusion, apart from the improbability of so great an effort at realism, is that the dialects used for instance even in the Mṛcchakaṭikā are clearly literary and not attempts to reproduce true vernaculars.

The stage reached by the Prākrits of Açvaghoṣa shows clearly how late are the Prākrits of the orthodox classical drama, [146] and reminds us how much more closely akin to Sanskrit must have been the Prākrit of the drama of the time of, or shortly after, Patañjali. The classical drama with its broken-down forms of Prākrit gives a false impression of the original dramatic form in which either perhaps Sanskrit alone, if the matter were epic, or both Sanskrit and a closely akin Çaurasenī appeared.

8. THE LITERARY ANTECEDENTS OF THE DRAMA

The drama owes in part its origin to the epics of India; from them throughout its history it derives largely its inspiration, far more truly so indeed than Greek tragedy as compared with the Greek epic. [147] From the epics also developed the Kāvya, the refined and polished epic, which appears at its best in the Kumārasambhava and Raghuvaṅça of Kālidāsa. The parallelism between the developed form of both is close and striking. The Sāhityadarpaṇa [148] lays down that it is a composition in several cantos, the hero a god or Kṣatriya of high race, of the type noble and superior; if there are several heroes, they are persons of royal rank of one family. The sentiment which predominates is the erotic, the heroic, or occasionally that of calm; the others serve in a subsidiary rôle. The subject-matter is either taken from tradition or not, but the heroes must be virtuous. The work begins with a prayer, a benediction, or an indication of the subject-matter. The development of the story employs the same five junctures as the theory prescribes for the drama. One or other of the four aims of man, wealth, love, performance of duty, or release, is to be attained by the action. The number of cantos is not to be less than eight; each should end in a different metre, and should announce the subject of the following act. Descriptions of every kind are essential; objects of these are the different times of day, the sun, the moon, night, the dawn, twilight, darkness, morning, midday, the hunt, mountains, the seasons, forests, the ocean, the sky, a town, the pleasures of love, the misery of separation from one’s beloved, a sacrifice, a battle, the march of an army, a marriage, the birth of a son, all of which should be developed in appropriate detail.

The essential feature of these little epics is the enormous development of the art of description, and the feature occurs in the other forms of narrative literature, the Kathā, tale, and the Ākhyāyikā, romance, types which blend with each other. Whether the subject be an imaginary theme, as is the Vāsavadattā of Subandhu, or a historical one, as in the Harṣacarita of Bāṇa, we find nothing treated as really important save the descriptions as contrasted with the narrative. The Sanskrit lyric also, in Kālidāsa’s masterpiece, the Meghadūta, is essentially descriptive, as is the Prākrit lyric preserved in the collection of Hāla, which is based on the model of an older lyric in Sanskrit, whose existence is revealed to us by the Mahābhāṣya.

The love of description, however, is not new; it is a characteristic of the epic itself, and the Rāmāyaṇa in special shows us how the way for the court poetry was being prepared. [149] Hence the fact that the verses of the drama are overwhelmingly descriptive, when not gnomic in character, is no matter for surprise. The peculiarity is a direct inheritance from the epic.

This fact has one important bearing on the history of the drama. The suggestion of Pischel [150] that the verses alone were once preserved, and the prose left to be improvised would have been plausible only if the verses had been essentially the important elements in the dialogue, as in the supposed Vedic Ākhyāna hymns. But this is assuredly not the case; the verses do little to help on the action; as in the epic, they express descriptions of situations and emotions; when movement of the play is requisite recourse is had to prose. Or the verses serve to set out maxims, as is natural in view of the great fondness of India for gnomic poetry, seen already in the verses introduced into the legend of Çunaḥçepa in the Aitareya Brāhmaṇa. In this again there is a close parallel with the epic, nor is it surprising that the epic poet, like Açvaghoṣa and Kālidāsa, was often devoted to the drama.

A further source of literary inspiration must undoubtedly be seen in the work of the lyric poets, of whose work clear evidence, as well as some scattered fragments, is preserved to us in the Mahābhāṣya of Patañjali. [151] Moreover, to these lyric writers it is probable that the drama owed some of its metrical variety; in the development of the metres with a fixed number of syllables, each of determined length, from the older and freer Vedic and epic forms, it may be taken as certain that the erotic poets, who had a narrow theme to handle, and had every motive to aim at variety of form and effect, must have contributed largely, a conclusion which is also strongly suggested, if not proved, by the very names of the metres with their erotic suggestion. [152]