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 [73]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 [74]name of Hāla, perhaps to be ascribed to the third or fifth century A.D.106

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-Māgadhī107 he knew, and the fact that the speaker of Old Māgadhī is the Duṣṭa, or bad man, reminds us of the bad character enjoyed108 by the Magadha. Lévi’s109 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 Grierson110 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. [75]

The stage reached by the Prākrits of Açvaghoṣa shows clearly how late are the Prākrits of the orthodox classical drama,111 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.