The real history of the Kavya, or artificial epic poetry of India, does not begin till the first half of the seventh century A.D., with the reign of King Harsha-vardhana of ThaneÇar and Kanauj (606–648), who ruled over the whole of Northern India, and under whose patronage Ba?a wrote his historical romance, Harsha-charita, and other works. The date of no Kavya before this landmark has as yet been fixed with certainty. One work, however, which is dominated by the Kavya style, the B?ihatsam?hita of the astronomer Varahamihira, can without hesitation be assigned to the middle of the sixth century. But as to the date of the most famous classical poets, Kalidasa, Subandhu, Bharavi, Gu?a?hya, and others, we have no historical authority. The most definite statement that can be made about them is that their fame was widely diffused by about 600 A.D., as is attested by the way in which their names are mentioned in Ba?a and in an inscription of 634 A.D. Some of them, moreover, like Gu?a?hya, to whose work Subandhu repeatedly alludes, must certainly belong to a much earlier time. The scanty materials supplied by the poets themselves, which might help to determine their dates, are difficult to utilise, because the history of India, both political and social, during the first five centuries of our era, is still involved in obscurity.

With regard to the age of court poetry in general, we have the important literary evidence of the quotations in Patanjali’s Mahabhashya, which show that Kavya flourished in his day, and must have been developed before the beginning of our era. Several of these quoted verses are composed in the artificial metres of the classical poetry, while the heroic anush?ubh Çlokas agree in matter as well as form, not with the popular, but with the court epics.

We further know that AÇvaghosha’s Buddha-charita, or “Doings of Buddha,” was translated into Chinese between 414 and 421 A.D. This work not only calls itself a mahakavya, or “great court epic,” but is actually written in the Kavya style. AÇvaghosha was, according to the Buddhist tradition, a contemporary of King Kanishka, and would thus belong to the first century A.D. In any case, it is evident that his poem could not have been composed later than between 350 and 400 A.D. The mere fact, too, that a Buddhist monk thus early conceived the plan of writing the legend of Buddha according to the rules of the classical Sanskrit epic shows how popular the Brahmanical artificial poetry must have become, at any rate by the fourth century A.D., and probably long before.

The progress of epigraphic research during the last quarter of a century has begun to shed considerable light on the history of court poetry during the dark age embracing the first five centuries of our era. Mr. Fleet’s third volume of the Corpus Inscriptionum Indicarum contains no fewer than eighteen inscriptions of importance in this respect. These are written mostly in verse, but partly also in elevated prose. They cover a period of two centuries, from about 350 to 550 A.D. Most of them employ the Gupta era, beginning A.D. 319, and first used by Chandragupta II., named Vikramaditya, whose inscriptions and coins range from A.D. 400 to 413. A few of them employ the Malava era, the earlier name of the Vikrama era, which dates from 57 B.C. Several of these inscriptions are praÇastis or panegyrics on kings. An examination of them proves that the poetical style prevailing in the fourth, fifth, and sixth centuries did not differ from that of the classical Kavyas which have been preserved. Samudragupta, the second of the Gupta line, who belongs to the second half of the fourth century, was, we learn, himself a poet, as well as a supporter of poets. Among the latter was at least one, by name Harishe?a, who in his panegyric on his royal patron, which consists of some thirty lines (nine stanzas) of poetry and about an equal number of lines of prose, shows a mastery of style rivalling that of Kalidasa and Da??in. In agreement with the rule of all the Sanskrit treatises on poetics, his prose is full of inordinately long compounds, one of them containing more than 120 syllables. In his poetry he, like Kalidasa and others, follows the Vidarbha style, in which the avoidance of long compounds is a leading characteristic. In this style, which must have been fully developed by A.D. 300, is also written an inscription by Virasena, the minister of Chandragupta II., Samudragupta’s successor.

A very important inscription dates from the year 529 of the Malava (Vikrama) era, or A.D. 473. It consists of a poem of no fewer than forty-four stanzas (containing 150 metrical lines), composed by a poet named Vatsabha??i, to commemorate the consecration of a temple of the sun at DaÇapura (now Mandasor). A detailed examination of this inscription not only leads to the conclusion that in the fifth century a rich Kavya literature must have existed, but in particular shows that the poem has several affinities with Kalidasa’s writings. The latter fact renders it probable that Vatsabha??i, a man of inferior poetic talent, who professes to have produced his work with effort, knew and utilised the poems of Kalidasa. The reign of Chandragupta Vikramaditya II., at the beginning of the fifth century A.D., therefore seems in the meantime the most probable approximate date for India’s greatest poet.

Besides the epigraphic evidence of the Gupta period, we have two important literary prose inscriptions of considerable length, one from Girnar and the other from Nasik, both belonging to the second century A.D. They show that even then there existed a prose Kavya style which, in general character and in many details, resembled that of the classical tales and romances. For they not only employ long and frequent compounds, but also the ornaments of alliteration and various kinds of simile and metaphor. Their use of poetical figures is, however, much less frequent and elaborate, occasionally not going beyond the simplicity of the popular epic. They are altogether less artificial than the prose parts of Harishe?a’s Kavya, and À fortiori than the works of Da??in. Subandhu, and Ba?a. From the Girnar inscription it appears that its author must have been acquainted with a theory of poetics, that metrical Kavyas conforming to the rules of the Vidarbha style were composed in his day, and that poetry of this kind was cultivated at the courts of princes then as in later times. It cannot be supposed that Kavya literature was a new invention of the second century; it must, on the contrary, have passed through a lengthened development before that time. Thus epigraphy not merely confirms the evidence of the Mahabhashya that artificial court poetry originated before the commencement of our era, but shows that that poetry continued to be cultivated throughout the succeeding centuries.

These results of the researches of the late Professor BÜhler and of Mr. Fleet render untenable Professor Max MÜller’s well-known theory of the renaissance of Sanskrit literature in the sixth century, which was set forth by that scholar with his usual brilliance in India, what can it Teach us? and which held the field for several years.

Professor Max MÜller’s preliminary assertion that the Indians, in consequence of the incursions of the Çakas (Scythians) and other foreigners, ceased from literary activity during the first two centuries A.D., is refuted by the evidence of the last two inscriptions mentioned above. Any such interruption of intellectual life during that period is, even apart from epigraphical testimony, rendered highly improbable by other considerations. The Scythians, in the first place, permanently subjugated only about one-fifth of India; for their dominion, which does not appear to have extended farther east than Mathura (Muttra), was limited to the Panjab, Sindh, Gujarat, Rajputana, and the Central Indian Agency. The conquerors, moreover, rapidly became Hinduised. Most of them already had Indian names in the second generation. One of them, Ushabhadata (the Sanskrit ?ishabhadatta), described his exploits in an inscription composed in a mixture of Sanskrit and Prakrit. Kanishka himself (78 A.D.), as well as his successors, was a patron of Buddhism; and national Indian architecture and sculpture attained a high development at Mathura under these rulers. When the invaders thus rapidly acquired the civilisation of the comparatively small portion of India they conquered, there is no reason to assume the suppression of literary activity in that part of the country, much less in India as a whole.

The main thesis of Professor Max MÜller is, that in the middle of the sixth century A.D. the reign of a King Vikramaditya of Ujjain, with whom tradition connected the names of Kalidasa and other distinguished authors, was the golden age of Indian court poetry. This renaissance theory is based on Fergusson’s ingenious chronological hypothesis that a supposed King Vikrama of Ujjain, having expelled the Scythians from India, in commemoration of his victory founded the Vikrama era in 544 A.D., dating its commencement back 600 years to 57 B.C. The epigraphical researches of Mr. Fleet have destroyed Fergusson’s hypothesis. From these researches it results that the Vikrama era of 57 B.C., far from having been founded in 544 A.D., had already been in use for more than a century previously under the name of the Malava era (which came to be called the Vikrama era about 800 A.D.). It further appears that no Çakas (Scythians) could have been driven out of Western India in the middle of the sixth century, because that country had already been conquered by the Guptas more than a hundred years before. Lastly, it turns out that, though other foreign conquerors, the Hu?as, were actually expelled from Western India in the first half of the sixth century, they were driven out, not by a Vikramaditya, but by a king named YaÇodharman Vish?uvardhana.

Thus the great King Vikramaditya vanishes from the historical ground of the sixth century into the realm of myth. With Vikramaditya an often-quoted but ill-authenticated verse occurring in a work of the sixteenth century associates Dhanvantari, Kshapa?aka, Amarasim?ha, Varahamihira, and Vararuchi as among the “nine gems” of his court. With the disappearance of Vikrama from the sixth century A.D. this verse has lost all chronological validity with reference to the date of the authors it enumerates; it is even inadmissible to conclude from such legendary testimony that they were contemporaries. Even though one of them, Varahamihira, actually does belong to the sixth century, each of them can now only be placed in the sixth century separately and by other arguments. Apart from the mythical Vikramaditya, there is now no reason to suppose that court poetry attained a special development in that century, for Harishe?a’s paneygyric, and some other epigraphic poems of the Gupta period, show that it flourished greatly at least two hundred years earlier.

None of the other arguments by which it has been attempted to place Kalidasa separately in the sixth century have any cogency. One of the chief of these is derived from the explanation given by the fourteenth-century commentator, Mallinatha, of the word dignaga, “world-elephant,” occurring in the 14th stanza of Kalidasa’s Meghaduta. He sees in it a punning allusion to Dignaga, a hated rival of the poet. This explanation, to begin with, is extremely dubious in itself. Then it is uncertain whether Mallinatha means the Buddhist teacher Dignaga. Thirdly, little weight can be attached to the Buddhistic tradition that Dignaga was a pupil of Vasubandhu, for this statement is not found till the sixteenth century. Fourthly, the assertion that Vasubandhu belongs to the sixth century depends chiefly on the Vikramaditya theory, and is opposed to Chinese evidence, which indicates that works of Vasubandhu were translated in A.D. 404. Thus every link in the chain of this argument is very weak.

The other main argument is that Kalidasa must have lived after Aryabha?a (A.D. 499), because he shows a knowledge of the scientific astronomy borrowed from the Greeks. But it has been shown by Dr. Thibaut that an Indian astronomical treatise, undoubtedly written under Greek influence, the Romaka Siddhanta, is older than Aryabha?a, and cannot be placed later than A.D. 400. It may be added that a passage of Kalidasa’s Raghuvam?Ça (xiv. 40) has been erroneously adduced in support of the astronomical argument, as implying that eclipses of the moon are due to the shadow of the earth: it really refers only to the spots in the moon as caused, in accordance with the doctrine of the Pura?as, by a reflection of the earth.

Thus there is, in the present state of our knowledge, good reason to suppose that Kalidasa lived not in the sixth, but in the beginning of the fifth century A.D. The question of his age, however, is not likely to be definitely solved till the language, the style, and the poetical technique of each of his works have been minutely investigated, in comparison with datable epigraphic documents, as well as with the rules given by the oldest Sanskrit treatises on poetics.

As the popular epic poetry of the Mahabharata was the chief source of the Pura?as, so the Ramaya?a, the earliest artificial epic, was succeeded, though after a long interval of time, by a number of Kavyas ranging from the fifth to the twelfth century. While in the old epic poetry form is subordinated to matter, it is of primary importance in the Kavyas, the matter becoming more and more merely a means for the display of tricks of style. The later the author of a Kavya is, the more he seeks to win the admiration of his audience by the cleverness of his conceits and the ingenuity of his diction, appealing always to the head rather than the heart. Even the very best of the Kavyas were composed in more strict conformity, with fixed rules than the poetry of any other country. For not only is the language dominated by the grammatical rules of Pa?ini, but the style is regulated by the elaborate laws about various forms of alliteration and figures of speech laid down in the treatises on poetics.

The two most important Kavyas are Kalidasa’s Raghuvam?Ça and Kumara-sambhava, both distinguished by independence of treatment as well as considerable poetical beauty. They have several stanzas in common, many others which offer but slight variations, and a large number of passages which, though differing in expression, are strikingly analogous in thought. In both poems, too, the same metre is employed to describe the same situation. In both poems each canto is, as a rule, composed in one metre, but changes with the beginning of the new canto. The prevailing metres are the classical form of the anush?ubh and the upajati, a development of the Vedic trish?ubh.

The Raghuvam?Ça, or “Race of Raghu,” which consists of nineteen cantos, describes the life of Rama together with an account of his forefathers and successors. The first nine cantos deal with his nearest four ancestors, beginning with Dilipa and his son Raghu. The story of Rama occupies the next six (x.–xv.), and agrees pretty closely with that in the Ramaya?a of Valmiki, whom Kalidasa here (xv. 41) speaks of as “the first poet.” The following two cantos are concerned with the three nearest descendants of Rama, while the last two run through the remainder of twenty-four kings who reigned in Ayodhya as his descendants, ending rather abruptly with the death of the voluptuous King Agnivar?a. The names of these successors of Rama agree closely with those in the list given in the Vish?u-pura?a.

The narrative in the Raghuvam?Ça moves with some rapidity, not being too much impeded by long descriptions. It abounds with apt and striking similes and contains much genuine poetry, while the style, for a Kavya, is simple, though many passages are undoubtedly too artificial for the European taste. The following stanza, sung by a bard whose duty it is to waken the king in the morning (v. 75), may serve as a specimen—

More than twenty commentaries on the Raghuvam?Ça are known. The most famous is the Sam?jivani of Mallinatha, who explains every word of the text, and who has the great merit of endeavouring to find out and preserve the readings of the poet himself. He knew a number of earlier commentaries, among which he names with approval those of Dakshi?avarta and Natha. The latter no longer exist. Among the other extant commentaries may be mentioned the Subodhini, composed by Dinakara MiÇra in 1385, and the ÇiÇuhitaishi?i, by a Jain named Charitravardhana, of which Dinakara’s work appears to be an epitome.

The Kumara-sambhava, or the “Birth of the War-god,” consists, when complete, of seventeen cantos. The first seven are entirely devoted to the courtship and wedding of the god Çiva and of Parvati, daughter of Himalaya, the parents of the youthful god. This fact in itself indicates that description is the prevailing characteristic of the poem. It abounds in that poetical miniature painting in which lies the chief literary strength of the Indian. Affording the poet free scope for the indulgence of his rich and original imaginative powers, it is conspicuous for wealth of illustration. The following rendering of a stanza in the Viyogini metre (in which lines of ten and eleven syllables ending iambically alternate) may serve as a specimen. The poet shows how the duty of a wife following her husband in death is exemplified even by objects in Nature poetically conceived as spouses—

Usually the first seven cantos only are to be found in the printed editions, owing to the excessively erotic character of the remaining ten. The poem concludes with an account of the destruction of the demon Taraka, the object for which the god of war was born.

More than twenty commentaries on the Kumara-sambhava have been preserved. Several of them are by the same authors, notably Mallinatha, as those on the Raghuvam?Ça.

The subject-matter of the later Kavyas, which is derived from the two great epics, becomes more and more mixed up with lyric, erotic, and didactic elements. It is increasingly regarded as a means for the display of elaborate conceits, till at last nothing remains but bombast and verbal jugglery. The Bha??i-kavya, written in Valabhi under King Çridharasena, probably in the seventh century, and ascribed by various commentators to the poet and grammarian Bhart?ihari (died 651 A.D.), deals in 22 cantos with the story of Rama, but only with the object of illustrating the forms of Sanskrit grammar.

The Kiratarjuniya describes, in eighteen cantos, the combat, first narrated in the Mahabharata, between Çiva, in the guise of a Kirata or mountaineer, and Arjuna. It cannot have been composed later than the sixth century, as its author, Bharavi, is mentioned in an inscription of 634 A.D. The fifteenth canto of this poem contains a number of stanzas illustrating all kinds of verbal tricks like those described in Da??in’s KavyadarÇa. Thus one stanza (14) contains no consonant but n (excepting a t at the end);1 while each half-line in a subsequent one (25), if its syllables be read backwards, is identical with the other half.2

The ÇiÇupala-vadha, or “Death of ÇiÇupala,” describes, in twenty cantos, how that prince, son of a king of Chedi, and cousin of K?ish?a, was slain by Vish?u. Having been composed by the poet Magha, it also goes by the name of Magha-kavya. It probably dates from the ninth, and must undoubtedly have been composed before the end of the tenth century. The nineteenth canto is full of metrical puzzles, some of a highly complex character (e.g. 29). It contains an example of a stanza (34) which, if read backwards, is identical with the preceding one read in the ordinary way. At the same time this Kavya is, as a whole, by no means lacking in poetical beauties and striking thoughts.

The Naishadhiya (also called Naishadha-charita), in twenty-two cantos, deals with the story of Nala, king of Nishada, the well-known episode of the Mahabharata. It was composed by Çriharsha, who belongs to the latter half of the twelfth century.

These six artificial epics are recognised as Mahakavyas, or “Great Poems,” and have all been commented on by Mallinatha. The characteristics of this higher class are set forth by Da??in in his KavyadarÇa, or “Mirror of Poetry” (i. 14–19). Their subjects must be derived from epic story (itihasa), they should be extensive, and ought to be embellished with descriptions of cities, seas, mountains, seasons, sunrise, weddings, battles fought by the hero, and so forth.

An extensive Mahakavya, in fifty cantos, is the Haravijaya, or “Victory of Çiva,” by a Kashmirian poet named Ratnakara, who belongs to the ninth century.

Another late epic, narrating the fortunes of the same hero as the Naishadhiya, is the Nalodaya, or “Rise of Nala,” which describes the restoration to power of King Nala after he had lost his all. Though attributed to Kalidasa, it is unmistakably the product of a much later age. The chief aim of the author is to show off his skill in the manipulation of the most varied and artificial metres, as well as all the elaborate tricks of style exhibited in the latest Kavyas. Rhyme even is introduced, and that, too, not only at the end of, but within metrical lines. The really epic material is but scantily treated, narrative making way for long descriptions and lyrical effusions. Thus the second and longest of the four cantos of the poem is purely lyrical, describing only the bliss of the newly-wedded pair, with all kinds of irrelevant additions.

The culmination of artificiality is attained by the Raghava-pa??aviya, a poem composed by Kaviraja, who perhaps flourished about A.D. 800. It celebrates simultaneously the actions of Raghava or Rama and of the Pa??ava princes. The composition is so arranged that by the use of ambiguous words and phrases the story of the Ramaya?a and the Mahabharata is told at one and the same time. The same words, according to the sense in which they are understood, narrate the events of each epic. A tour de force of this kind is doubtless unique in the literatures of the world. Kaviraja has, however, found imitators in India itself.

A Mahakavya which is as yet only known in MS. is the Navasahasanka-charita, a poem celebrating the doings of Navasahasanka, otherwise Sindhuraja, a king of Malava, and composed by a poet named Padmagupta, who lived about 1000 A.D. It consists of eighteen cantos, containing over 1500 stanzas in nineteen different metres. The poet refrains from the employment of metrical tricks; but he greatly impedes the progress of the narrative by introducing interminable speeches and long-winded descriptions.

We may mention, in conclusion, that there is also an epic in Prakrit which is attributed to Kalidasa. This is the Setu-bandha, “Building of the Bridge,” or Rava?avadha, “Death of Rava?a,” which relates the story of Rama. It is supposed to have been composed by the poet to commemorate the building of a bridge of boats across the Vitasta (Jhelum) by King Pravarasena of Kashmir.

There are a few prose romances dating from the sixth and seventh centuries, which being classed as Kavyas by the Sanskrit writers on poetics, may be mentioned in this place. The abundant use of immense compounds, which of course makes them very difficult reading, is an essential characteristic of the style of these works. As to their matter, they contain but little action, consisting largely of scenes which are strung together by a meagre thread of narrative, and are made the occasion of lengthy descriptions full of long strings of comparisons and often teeming with puns. In spite, however, of their highly artificial and involved style, many really poetical thoughts may be found embedded in what to the European taste is an unattractive setting.

The DaÇa-kumara-charita, or “Adventures of the Ten Princes,” contains stories of common life and reflects a corrupt state of society. It is by Da??in, and probably dates from the sixth century A.D. Vasavadatta, by Subandhu, relates the popular story of the heroine Vasavadatta, princess of Ujjayini, and Udayana, king of Vatsa. It was probably written quite at the beginning of the seventh century. Slightly later is Ba?a’s Kadambari, a poetical romance narrating the fortunes of a princess of that name. Another work of a somewhat similar character by the same author is the Harsha-charita, a romance in eight chapters, in which Ba?a attempts to give some account of the life of King Harshavardhana of Kanauj. There is, however, but little narrative. Thus in twenty-five pages of the eighth chapter there are to be found five long descriptions, extending on the average to two pages, to say nothing of shorter ones. There is, for instance, a long disquisition, covering four pages, and full of strings of comparisons, about the miseries of servitude. A servant, “like a painted bow, is for ever bent in the one act of distending a string of imaginary virtues, but there is no force in him; like a heap of dust-sweepings gathered by a broom, he carries off toilet-leavings; like the meal offered to the Divine Mothers, he is cast out into space even at night; like a pumping machine, he has left all weight behind him and bends even for water,” and so on. Soon after comes a description, covering two pages, of the trees in a forest. This is immediately followed by another page enumerating the various kinds of students thronging the wood in order to avail themselves of the teaching of a great Buddhist sage; they even include monkeys busily engaged in ritual ceremonies, devout parrots expounding a Buddhist dictionary, owls lecturing on the various births of Buddha, and tigers who have given up eating flesh under the calming influence of Buddhist teaching. Next comes a page describing the sage himself. “He was clad in a very soft red cloth, as if he were the eastern quarter of the sky bathed in the morning sunshine, teaching the other quarters to assume the red Buddhist attire, while they were flushed with the pure red glow of his body like a ruby freshly cut.” Soon after comes a long account, bristling with puns, of a disconsolate princess lying prostrate in the wood—“lost in the forest and in thought, bent upon death and the root of a tree, fallen upon calamity and her nurse’s bosom, parted from her husband and happiness, burned with the fierce sunshine and the woes of widowhood, her mouth closed with silence as well as by her hand, and held fast by her companions as well as by grief. I saw her with her kindred and her graces all gone, her ears and her soul left bare, her ornaments and her aims abandoned, her bracelets and her hopes broken, her companions and the needle-like grass-spears clinging round her feet, her eye and her beloved fixed within her bosom, her sighs and her hair long, her limbs and her merits exhausted, her aged attendants and her streaming tears falling down at her feet,” and so forth.