Impressive proof for the above links between the history of Brahman migration and the textual history of the SR of the Mahābhārata is furnished by the history of the Brāhmī scripts and their various derivatives, as it has been re-constructed by Iravatam Mahadevan (2003). We must keep in mind we cannot have a textual tradition without a phonologically appropriate script, linking, in other words, the epic to the relevant human agency, the third correlate in the equation. I begin with Mahadevan’s master chart for the entire development:
- 3rd Century BCE Brāhmī
- 2nd Century BCE Southern Brāhmī, Tamil Brāhmī
- 1st Century BCE Bhattripolu
- 5th Century CE Vaṭṭeḻuttu (from Tamil Brāhmī )
- 6th Century CE Proto-Telugu and Kannada, Grantha (from Southern Brāhmī)
- 7th Century CE Telugu, Kannada
- 7th Century CE Tamileḻuttu (from grantha and Vaṭṭeḻuttu)
- 14th Century Malayalam-Āryeḻuttu (from grantha and Vaṭṭeḻuttu)
We see that the Brāhmī script devolves into two separate and independent lines of developments, starting with the Southern Brāhmī and Tamil Brāhmī, arriving in peninsular India separately and giving rise to the five major historical scripts of the area, Telugu, Kannada, Grantha, on the one hand, and Tamil and Malayalam, on the other. The Southern Brāhmī script is seen to give rise to the first three, the Kannada and Telugu scripts emerging from an intermediate proto-script of the parent Southern Brāhmī and the Grantha, more directly from it. This latter fact has great significance for us. On the other hand, the Tamil Brāhmī script is seen first to evolve into Vaṭṭeḻuttu, which from reaction with the Southern Brāhmī derivative, Grantha, gives us the Āryeḻuttu script of Malayalam and Tamileḻuttu script of Tamil, (the latter, as we will see below but not shown in Mahadevan’s chart, showing a further influence of a Northern Brāhmī script—what we may call the σ-script after Sukthankar’s use of the Greek letter for the NR text that comes south with the Aparaśikhā Brahmans, from about 8th century CE, the period of the Aparaśikhā migration.)
These paleographical facts have significant bearing on the arguments presented above on the different genealogies of the Mahābhārata epic and their agents of transmission, the Brahman groups, that came to the peninsular India, starting with the Pūrvaśikhā Brahmans, arriving in the Tamil country well enough in time to take part in the production of the poetries of the Sangam period, and the Aparaśikhā Brahmans, arriving almost half a millennium afterward, under the Pallava patronage, from 5th century CE.
It is useful to consider the problem in its three main aspects:
- i. Introduction and an over-view of Mahadevan’s findings
- ii. the Tamil Brāhmī script and its history
- iii. the Southern Brāhmī script and its history
- iv. the Brahmans, the epics and paleography
D. i. Introduction and an Over-view of Mahadevan’s Findings
As Mahadevan (2003: 315) shows, the Tamil Brāhmī script is attested in the 3rd century BCE Jain cave inscriptions, starting with those of the Māṅgulam caves, around Madurai in the Pāṇṭiyan territory, the Pāṇṭiyan kings being thus the earliest and in the early period the most frequent hosts and patrons to the Jain monks and the Jain religion. 72 It is quite likely that the indigenous Tamil society at this time was largely oral, as Hart (1975:157) has argued, still in the phase of the pāṇan songs and their oral traditions and the latter in the process of beginning to become the templates for the literate and decidedly literary overlays of the Sangam songs, as they have come down to us. The Tamil Brāhmī script evolves over the next four centuries, providing the script for the Sangam-era compositions, dating from ca. 50 BCE to 200 CE, transforming into an early form of the Vaṭṭeḻuttu script by ca. 6th century CE and mature Vaṭṭeḻuttu script afterward. Correspondingly, the language itself changes from Old Tamil (250 BCE to 100 CE), represented by Tolkāppiyam and probably some Puṛanānūṛu songs, to middle Old Tamil (100 to 400 CE), represented by bardic poems on love and war collected in the Eṭṭutokai and Pattuppāṭṭu anthologies, into Late Old Tamil, (400-700 CE) with the two epics, Cilappatikāram and Manimēkalai, as its representative texts (Lehman 1994; Takahashi 1995; Steever 2004). The key point to note here is that there is a complete fit between Tamil phonology and Tamil Brāhmī script, and the body of Saṅgam, “academy” literature, cited so from the 7th century onward to signify the canon of the academy, cāṇṛor ceyyuḷ, “poetry of the nobles” (Steever 2004: 1037), runs into some 32,000 lines (Lehman 1998: 75).
The Southern Brāhmī script constitutes, on the other hand, an independent derivation from the parent Brāhmī script (Mahadevan 2003: 176), arising at the same time as the Tamil Brāhmī script, but it provides an entirely different history. The modern languages of Kannada and Telugu are the outcome at one line of development, thus through the western areas of the peninsular regions, but it gives rise to the Grantha script in the eastern parts, in the Toṇṭaimaṇṭalam region, appearing in epigraphy ca. 6th century CE, with what is considered to be the first Grantha inscription (213). We do not have much information in Mahadevan about their parallel evolutions other than that, at its attestation, the Tamil Brāhmī script is already the entrenched script of the Tamil country, fashioned, as Mahadevan argues, in the Jain monasteries around Madurai in the Pāṇṭiyan kingdom, ca. 3rd century BCE, already adapted to meeting the requirements of the Tamil phonology. As noted, this is the script in which the literate— and literary–overlay of the Sangam songs on the Pāṇan oral templates by the pulavan (“learned”) poets takes place (Hart 1975). On the other hand, the Southern Brāhmī script is attested along an independent line of descent in its Grantha form only ca. 6th century CE (Mahadevan: 213), meeting, it should be noted, the needs of the Sanskrit phonology. And in Mahadevan’s scheme, the Telugu and Kannada scripts are cohorts in this development.
We notice a gap of almost 600 years between the attestations of the two scripts in the Tamil country, the Tamil Brāhmī script by 250 BCE and the Southern Brāhmī script by 6th century CE, the first meeting Tamil phonology and the second meeting, the Sanskrit phonology. Because of the efflorescence of the Saṅgam poetry in this period of 600 years—largely in Old Middle Tamil and in Tamil Brāhmī script–we do not raise the question if there was literary activity in the peninsular region in Sanskrit in the same time period. We have already noted that a substantial number of these poets of Saṅgam poetry were Brahmans, wearing the pūrvaśikhā and using the Tamil Brāhmī syllabary to compose the songs. Was there no composition among them simultaneously in Sanskrit? And if so what script served them? These questions lead in turn to a fundamental question: if the Jains brought with them a script (the parent Tamil Brāhmī script) with them, did the Brahmans bring with them a script?
Yet this question is never posed. Consider for instance this statement by Lehman (1998:75), “During this period [Sangam], with the propagation of Jainism and Buddhism in South India a number of Prakrit and Sanskrit borrowing entered Old Tamil and appear in Sangam anthologies (my parenthetical gloss).” The arrival of Brahmanism is not similarly posed as an alien influence, presumably because the later Hinduism subsumes both Brahmans and non-Brahmans as one group in the Tamil country in contrast to the Buddhists and Jains. Yet for this period, Brahmanism in the form of its Śrauta ethos is just as alien in the cultural ecology of the Tamil country, and as Sangam poetry shows by far the most dominant. For instance, Mahadevan considers the presence of Buddhism in the Brāhmī inscriptions to be negligible, something that can be said with equal justice for its presence in Sangam anthologies as well. Jainism is the dominant religion in the inscriptions, but tapering off in time and almost totally eclipsed in Sangam literature. On the other hand, as we will see, the Brahman presence, just as alien in the context as the Jain and Buddhist, is on the ascendance. It is almost completely unattested in the Tamil- Brāhmī inscriptions, but as an alien presence, it dominates the Sangam anthologies: a good percentage of the Sangam poets are Brahmans; śrautism is decidedly extolled, a king coming to be named after the ritual hall where the sacrificial animal is immolated, the Pāṇṭiyan King, Paliayākacālai Muṭukuṭumip Peruvaḻuti. This poses a fundamental question to the recensional history of the epic: if the SR text arose as the *Pūrvaśikhā text in my chart in the first millennium of the CE, what script could have served the composition? We have placed the epic in the form of a *Śārada text and a human agency in the form of the Pūrvaśikhā Brahmans in the scene; we have now to place a script in the region, a script that can meet Sanskrit phonology. It is easy to see that the only option we have is the Southern Brāhmī derivative, the Grantha script. Thus, I would be arguing that the SR *Pūrvaśikhā text begins its life in a *Southern Brāhmī script, Grantha, or an early form of it, being the most logical candidate. Mahadevan (213) considers the Grantha script to be derived from Southern Brāhmī of the Prākṛt Charters of the Early Pallavas, 4-5th centuries CE. If my scenario that the SR rises in the first centuries of the CE, soon after the arrival of the Pūrvaśikhā Brahmans in the peninsular India with a *Śārada text of the epic is valid, the only script that can meet the demands of the literate composition of the SR is the Grantha script. I would be arguing below thus that a form of the Southern Brāhmī script, substantially similar or identical with this, arrived in the Tamil country with the Pūrvaśikhā Brahmans and was already present in the area when the Pallava reign begins. The attestation of the paviḻiya adherents, ca. 9th century CE, in the Toṇṭaimaṇṭalam area in the Pallava epigraphy, suggests that the Pūrvaśikhā Brahmans were present in this area as well, around Vēṇkata hills, after their dakṣināpatha migration. This is also the area of the Prākṛt Charters of the early Pallavas, which display the first epigraphic evidence of the Grantha script.
In sum, then, both the Tamil Brāhmī and the Southern Brāhmī scripts originate from a common parental *Brāhmī script (Mauryan?) and both are attested only in peninsular India, but at entirely different time intervals, the first by ca. 2rd century BCE and the second by only ca. 6th century CE. The Tamil Brāhmī script, eventually becoming the Vaṭṭeḻuttu of the Tamil-Kerala country, meets the linguistic needs of the Tamil language in the area, most significantly that of the Sangam poetry. On the other hand, the Southern Brāhmī scripts must be seen, in some incipient form of the later Grantha script, as the vehicle of the Southern Recension of the Mahābhārata, when it takes shape, in the first centuries of the Current Era in the same area.
D. ii. The Tamil Brāhmī Script
Based on Mahadevan’s chart given above, we can say that the Tamil Brāhmī arrived in South India in 3rd century BCE, and it was brought to peninsular India by the Jains, arriving there from the north, it is widely accepted, through Karnataka in the west and not through the Vēnkatam hills of the later Brahman migrations: it is likely, as Mahadevan (135) notes, that “Tamil Brāhmī script was adapted from the Mauryan Brāhmī in the Jain monasteries (‘paḷḷi’) of the Madurai regions sometime before the end of the third century BCE” (Mahadevan’s parenthesis). In the Early Period (3rd to 1st centuries BCE) in Mahadevan’s chronology, out of 30 sites with 86 Tamil-Brāhmī inscriptions, in Early Old Tamil, 28 sites with 84 inscriptions pertain to Jainism, and they are mostly in the Pāṇṭiyan region, around Madurai, leaving, as Mahadevan notes (128) “no longer any doubt that the Tamil-Brāhmī cave inscriptions are mostly associated with the Jaina faith.” In the Middle Period (1st to 3rd centuries CE), the period of the Middle Old Tamil, there is a sharp decline in cave inscriptions, and this is accompanied by a striking shift of Jainism from the Pāṇṭiyan kingdom to the Karur-based Cēra region, with the main trope of the inscriptional passages—the grant of the cave shelter to a Jain monk by a ruler—continuing, as for instance in the case of the Pugalūr site on the southern banks of the Kaveri river 15 kilometers northwest of Karur, dated to 3rd century CE (405- 421; Items 61 through 72). By Late Period (3rd to 5th centuries), that of the Late Old Tamil, the natural cave inscriptions come to an end, with the Sittanavasal B site (451- 461; Items 101 through 109), already in early Vaṭṭeḻuttu, being the last of the Jain cave shelters—giving way as well to a new kind of Jaina monuments in the form of nicītikai (← Kannada inscr. nisidige [Mahadevan: 632]) inscriptions, denoting a “seat of penance…where a Jaina monk performs the religious penance of fasting unto death” (Mahadevan: 632), the sallēkhana death (“death by starvation”) at Paṛaiyanpaṭṭu and Tirunātharkunru (470-473; #s 115 and 116 in Mahadevan’s numeration), ca. 6th century CE.
We are no longer in the oral society of the itinerant pāṇans now but in a fully literate period of Tamil history, the lasting legacy of Jainism, as Mahadevan (139) notes, to the Tamil history, leading to the efflorescence of the Sangam literature of the early centuries, CE.111 As Hart (1975) has conclusively argued, the Sangam poetry is a literate—and literary—copy created by a written overlay on the original oral templates of the pāṇan songs.112 The Tamil Brāhmī script gives us a script for this overlay, as indeed already suggested by Hart (147), the script in which these poems were written, presumably with an iron stylus on palm leaves, the stylus held in the tightly closed, ritually correct right fist, the technique and practice of the mode of writing, producing in time, presumably, the circular shape of the Vaṭṭeḻuttu script. We are at the juncture of the rise of the historical Tamil script, Tamil-eḻuttu, adapted, ca. 8th CE, from the Vaṭṭeḻuttu script and the Grantha script of the Southern Brāhmī filiation with as noted an input from a σ-script that came with the Aparaśikhā Brahmans: I come back to this in C. ii below.
We must note, however, that the Vaṭṭeḻuttu script remains, at this stage, in its pure and unalloyed form in the eastern and south-eastern parts of the Tamil country, as for instance in the famous Vēlvikkuṭi Plates of the 8th century, and covering besides most of the modern territory of Kerala.
It is striking that in this new literature of the Sangam poetry, written in a Jain- invented script, the Jains and Jainism are signally absent. Other than the solitary Akanānūru (123)113 reference to the Jain practice of sallēkhana death, the trope, as we saw, of the later, 6th CE, Late Period Tamil Brāhmī-Early Vaṭṭeḻuttu inscriptions— marking, it should be added, a Karnataka Jain practice, and not so much Tamil—aspects of Jainism itself are remarkably absent in the Sangam poetry.114 We do not have as yet an adequate explanation for this sudden decline of Jainism through the six centuries, from the Early Period (3rd to 1st centuries BCE) to the Middle Period (1st to 3rd centuries CE) and the Late Period, (3rd to 5th centuries CE). Why are the Jains and Jainism unrepresented or represented so meagerly in the Sangam poetry, generally accepted to be in composition in the first centuries of the Current Era?
Let us consider. The cave inscriptions testify to a deep and organized Jain establishment in the Tamil country from the 3rd century BCE onward. Mahadevan adduces (128-139) seven terms of various but precise significations for a Jain monk, from kaṇi (head of a gaṇa) through amaṇan (an ascetic), to upacaṇ (a lay teacher of scriptures) to māṇākkar, a student or novice. They appear linked to some 14 individual Jain names in these inscriptions: one Attiran (<Atri, a gotra term) is an amaṇan; Naṭṭi, Naṭan, Nākan, Nanda-Siri-Kuvan are kaṇis. We have seven dhārmic terms, like aṭittānam (< Skt. aṭisthāna), ‘seat’ of authority; aṛam, ‘charity or religious life’ and ‘paḷḷi,’ for hermitage, the last term also serving as the suffix in the names of many human settlements in the Tamil-Kerala country. Mahadevan (139) considers thus the contribution by the Jains to the Tamil history “enormous” and “most basic and fundamental”. The inscriptional evidence shows that the first stage in the decline of Jainism, or its royal patronage, is marked by the cessation of cave sites in the eastern parts of the Tamil country, the Pāṇṭiyan kingdom, and their shift to the west, in the Karur-based Cēra kingdom (the Pugalūr sites, Item XX: 1 through 12; Mahadevan: 405-421), later to produce landmark works by Jain authors, the Cilappatikāram and Cīvakacintāmaṇi, to name just two of the most noted texts. We must note as well that the inscriptional evidence points to continuous contacts between the Tamil Jains and the Jain centers of the Karnataka region, a point emphasized by Mahadevan (135).
It is useful to note that this is precisely the time period, the dawn of the Current Era, in which the Pūrvaśikhā Brahmans arrive in the Tamil country in the scheme presented above in A.i and to be taken up again in C. iii. below: they are clearly and concretely attested in the Sangam poetry with their pūrvaśikhā kuṭumi. Like the Jains, they also come from the north, but not through the Karnataka region, but through the dakṣiṇāpatha route in the lower Godavari region, possibly at Assaka in its banks, and further south through the Vēnkaṭa hills, and eventually into the kingdoms of the mūvēndar—the land of the three Indras, the Cēra, Coḷa, and Pāṇṭiya kings, the occurrence of the paviḻiya term in the Pallava epigraphy of the 8th century CE still placing them in the Toṇṭaimaṇṭalam area as late as 8th century CE.
We have already noted that the Vedic content of the Sangam poetry is considerable, and that a good 10% of the Sangam poets were Brahmans. We must add to this the evidence from the Sangam poetry that some of the foremost patrons of the Vedic 80 ritualism were the Pāṇṭiyan kings, erstwhile hosts to the Jain religion. Perhaps the most prominent of these kings is the great Paliyākacālai Muṭukuṭumip Peruvaḷuti, (of Puṛanānūṛu 6, 12, 15, 64)—such a patron of Vedic ritual as to be named after the yāgaśāla of the Vedic ritual, with the yūpa or the pole fixed just outside the eastern boundary of the ritual hall, on the pṛṣṭha axis, the line to the rising sun, to which the animal (‘bali’) is tethered to be sacrificed in a Soma class ritual. At Puṛanānūṛu 15. 11- 17, the poet-singer, Naṭṭimaiyār, almost certainly a Brahman, celebrates this king:
Given your fury, which of these is in greater in number
–your once eager enemies shamed and despairing after brandishing
their long spears that throw shadows and their beautiful shields
embossed with iron against the power of your swift vanguard
with its shining weapons, or else the number of spacious sites
where you have set up columns after performing many sacrifices
prescribed by the Four Vedas and the books of ritual
fine sacrifices of an excellence that will not die away[.]
Hart’s (2000) translation.
Yet Peru-vaḻuti’s namesake first appears in the Tamil-Brāhmī Māngulam I inscriptions, ca. 3rd century BCE, the oldest Tamil-Brāhmī inscription in the Pāṇṭiyan region and the oldest Jain inscription all of India, as “Kaṭalan Vaḻuti” (Vaḻuti of the Sea’), the paṇavan (“servant”) of Neṭuñceḻiyan, the Pāṇṭiyan king of the Māṅgulam I inscriptions, and who oversees the construction of the stone bed for the Jain kaṇi, Nanda-Siri-Kuvan (Mahadevan 2003: 315-323; Item I, 1 through 6). “Vaḻuti” is widely attested as a generic Pāṇṭiyan name, passing on later to Pāṇṭiyan kings—indeed, one of the two kings credited with the collection of two anthologies, Ainkuṛunūṛu and Akanānūṛu, being Ugra-pperu- vaḻuti. The Vaḻuti of the Māṅgulam I inscriptions need not thus be a direct ancestor of the later Muṭukuṭumip-Peruvaḻuti, the ‘big’ (peru) Vaḻuti, but the fall from favour of the Jains in the Pāṇṭiyan kingdom by the end of the Early Period (beginnings of the CE) of the Tamil Brāhmī paleography cannot be ignored. The first Vaḻuti is the paṇavan, the overseer of the construction of a stone bed for Nanda-Siri-Kuvan, the Jain kaṇi, whereas the “Big” Vaḻuti of the Sangam poetry, the patron of four of its songs, is seen to be synonymous with Vedic Śrautism, brought to the Tamil-Kerala country by the Pūrvaśikhā Brahmans. It is clear that the Brahmans of the Sangam period—that is, a period synchronous with the Middle Period of the Tamil- Brāhmī paleography, 1st to 3rd CE– replace the Jains of the Early Period of the Tamil Brāhmī paleography as the new recipients of royal patronage at the Pāṇṭiyan courts, with the Śrauta ritual, certainly more spectacular than the spectacles of the Jaina religion and more promising of worldly and other-worldly glory,115 forging the old Āryan brahma-kṣatra alliance between Brahmans and Kings, but now in the Tamil country, as the Rājasūya ritual of the Cōḷa king, Vēṭṭa Perunāṛkiḷḷi, shows. Indeed, as Hart notes (1975: 70-71), the Sangam poetry acknowledges, as at Puṛ.166, that “a struggle is under way between the orthodox and non- orthodox religions” with the Brahman (of the kauṇḍinya gotra) to whom the poem is addressed seen as establishing the truth “not agreeing with those who claim the true is false, and who realized the lie that seemed as if it were true to utterly defeat those who would quarrel with the one ancient book.” The śrauta ‘status kit’116 of the Brahmans wins the day, not for the first time, nor the last.
A corresponding Jain resentment at the Brahman usurpation of their patronage is not totally impossible, nor illogical, and only extreme political correctness, no doubt, a corrective reaction to the Brahman historiography of the Tamil country of the first five decades of the 20th century, would be blind to this.117 The continuous contact of the Tamil Jains with their Karnataka counterparts is an important element in this complex and changing picture. For, the next great historical event, and perhaps the most important in some ways of Tamil history as a whole, although not sufficiently understood, is the invasion of Tamil country by the Jain-Kaḷabhras from Karnataka, creating the famous Kaḷabhra Interregnum, the “long night” of the Tamil history in the extreme Brahman historiography of the subject, with the Pāṇtiyan kingdom receiving the brunt of the invasion.118 Thus while the Kaḷabhra’s anti-Brahmanical excesses may have been exaggerations of a Brahman historiography, there is wide-spread consensus that the Kaḷabhras were both Jain and from Karnataka, and their conquest and rule of the Tamil country over three centuries constituted a complete break with the classical Sangam period. As Mahadevan (136) notes, “[the Kaḷabhras] displaced the traditional Tamil monarchies and held sway over the Tamil country for nearly three centuries until they were expelled in the last quarter of 6th century CE by Kaṭunkōṇ, the Pāṇṭiya, from the south, and Simhaviṣṇu the Pallava from the north (my parenthesis).” It is an eighth descendant of this Kaṭuṇkōṇ, Neṭuñjaṭaiyan, who appears in the Vēḷvikkuṭi Plates (EI XVII (1923-24):271), restoring lands of the Vēḷvikkuṭi village to a Brahman petitioner by the name of Korkaikiḷan Nar Ciṅkan, originally gifted, as recorded in the plates, to his ancestor Korkaikiḷan Narkoṟṟan, by the great Paliyākacālai Muṭukuṭumip Peruvaḷuti of the Sangam poetry.
We have here a grid of three Pāṇṭian kings and three Brahman beneficiaries spread over some six centuries. Peruvaḷuti of the Sangam period (ca. 200 CE) gifts the village of Vēḷvikkuṭi to a Śrauta Brahman, Narkoṟṟan, the village acquiring its name from Tamil vēlvi (“sacrifice”) from Narkoṟṟtan’s śrauta ritual at the site; the Kaḷabhras dispossess his descendants of this gift some length of time later, perhaps two centuries. King Kaṭuṇkōn, in marking the end of the Kaḷabhra Interregnum, restores the Vēḷvikkuṭi land grant to an unnamed Brahman descendant of the original donee, ca. 620 CE, almost four centuries after the Peruvaḷuti grant of the Sangam period. All of this is ratified by King Neṭuñjaṭaiyan, the issuer of the Vēḷvikkuṭi Plates, seven kings after Kaṭuṇkōn, thus ca.760 CE, by affirming the right of Nar Ciṅkan, the petitioner and remote descendant, indeed, of the original donee, Narkoṟṟan. And Narkoṟṟan’s patron, King Peruvaḻuti of the Sangam period looms as the prime mover of the narrative, himself linked at least by name to a Vaḻuti of the Māṅgulam Plates and a patron of the Jains. We are thus witness to a period of Jain dominance and patronage, a Brahman usurpation of their patronage in the Pāṇtiyan court, a Jain disruption of the established order of the Tamil society through the Kaḷabhra Interregnum, and an eventual Brahman restoration.
I would suggest that part of the disruption of the Kaḷabhra period also results in the break-up of the first Brahman group of the Tamil country, the Pūrvaśikhā group, into its historical remnants. We first see them in the Tamil country in the Sangam poetry, portrayed in it with their kuṭumi in the likeness of a horse’s mane, composing themselves a sizeable number of these poems, no doubt using the Tamil Brāhmī script, created by the Jain monks in the Pāṇṭiyan kingdom almost two centuries before. After the Kaḷabhra Interregnum, we begin to see the Nambudiri Pūrvaśikhās in Malabar across the Palghat gaps facing the Karur-based Cēra kingdom, certainly the śrauta elites of the community, and, as Mahadevan notes, creating from 10th to 16th CE the historical Malayalam script from the Vaṭṭeḻuttu and Grantha script, called locally the Āryeḻuttu (2003: 212). However, Mahadevan does not explain how the Nambudiri Pūrvaśikhās come to possess the Grantha script, by 10th century CE. True, the Grantha script has already been in existence, but in the Tamil country proper, for almost half a millennium, and Mahadevan does not explain how it comes to the Nambudiris, in Kerala. It is unlikely that the Grantha script arrived in a disembodied form to Malabar and to the Nambudiris; it is equally unlikely that the conservative Nambudiris would have accepted a script from the outside. Besides, composition in Sanskrit went apace among the Nambudiri Pūrvaśikhās in Malabar before the 10th century CE, showing the presence of a Sanskrit-able script in the region. We must note too that almost all intercourse between the Tamil country and the emerging Kerala entity had ceased by the 10th century CE, Mahadevan’s date for the start of the formation of the Āryeḻuttu In my scheme, the script would have accompanied the Nambudiri Pūrvaśikhās at their departure at the Kaḷabhra Interregnum to the Malabar area: indeed, it is the script of *Pūrvaśikhā Mahābhārata, the archetypal Southern Recension text that was found in the Nambudiri houses and centers of learning in the 20th century: I consider this in fuller detail in Section D iii below.
D. iii. The Southern Brāhmī Script
This is the other script into which the Mauryan Brāhmī originally devolves and which, like its counterpart, the Tamil Brāhmī script, came to the Tamil-Kerala country, giving us three historical South Indian scripts, Kannada and Telugu on the one hand, by 85 6th to 7th centuries, and the Grantha script, on the other, a little earlier, by 5th CE. As we have already seen, Mahadevan has persuasively suggested that the Tamil Brāhmī script was fashioned by the Jain monks ca. 3rd BCE in Madurai Jain monasteries, and this script fashions the course of Tamil history for the next half a millennium, functioning as the script of the Sangam poetry and transforming later into the Vaṭṭeḻuttu script and serving vast areas of the Tamil-Kerala country, all along the east coast of the Tamil country and all of today’s Kerala. But what about the origins of its sister script, the Southern Brāhmī script, and its development? Who brought it to the south? Why was it not attested till ca. 5th century CE, with the first Grantha inscription, marking a 600-year gap between the Tamil Brāhmī derivatives and Southern Brāhmī derivatives?
Answers to these and related questions lie in the scenario I have been advancing regarding Brahman migration to the south—especially with the Mahābhārata epic. In fact, we will see that it is the epic half of the story that completes the validity of the argument presented above: the departure of the Pūrvaśikhā Brahmans, ca.150 BCE from the antarvedi area of the Ganga-Yamuna doab with a version of the epic resonant with the *Śārada text of the Mahābhārata epic and their arrival in the Tamil country in time to be attested in the Sangam poetry both as players in the poems and their composers on the one hand, and fashioning on the other hand, the *Pūrvaśikhā version of the Southern Recension in the half millennium or so after their arrival, by the Kaḷabhra Interregnum. The question that will elucidate the entire problem concerns the script in which the *Śārada text came to the south with the Pūrvaśikhā Brahmans. The Pūrvaśikhā Brahmans have displayed strong oral traditions; the famous example of the Nambudiri Pūrvaśikhās is only the most conspicuous one. As Raghavan notes in the 1958 survey of the state of all-India Vedic recitation, the Śōḻiya Brahmans also possess live family-based Vedic oral traditions.119 Something similar to this could be said about the two other temple-based Pūrvaśikhā groups as well, the Chidambaram Dīkṣitars and the Tiruchendur Mukkāṇi Brahmans, although outside the Vedic tradition properly so called.
With this in background, we could raise the question if the Pūrvaśikhā Brahmans brought the *Sārada text as an oral archive. The *Sarada text, as it has been assembled in the Poona CE, runs into 75,000 verses—not a formidable number for a person oriented and trained in the arts and sciences of the oral tradition to commit and transmit in a memorial tradition: we have the example of a Nambudiri Pūrvaśikhā, Ēṛkkara Rāman Nambudiri, dictating the entire text of the Kauṣītaki Brāhamaṇa from memory to E.R. Sreekrishna Sarma in 1968, rather to a tape recorder commandeered by Professor Sarma for the task of the textualization of the text.120 And this would have been only part of his oral repertory; as a Kauṣītaki Ṛgvedi, he would know by memory all of the Ṛgveda from the saṃhita mode to the jaṭa vikṛti as well as the Āraṇyaka and Upaniṣadic texts of his birth Veda, all part of the svādhyāya regimen of his family. The memory load of 75,000 verses is not the problem per se, inside the context of a fully functioning and flourishing system of oral tradition, as we know the Vedic system to have been. The problem lies in the fact, on the other hand, that there would have been no need nor use for the memorization of the epics, as no rituals demand intact recitation of verses from the epic as is the case with Vedic verses—ignoring for the moment the oral origins of the epic, the original oral pragmatics that gave rise to the epic at its formative stages. We must note that the various mnemonic devices associated with the Vedic oral traditions—the padapātha and its vikṛti modifications—possess no epic counterpart.
Verses from the epic never really possessed a ritual context, demanding the phonetically correct recitations, as we know was the case with the Vedic verses. In other words, there is no oral infrastructure for the transmission of the epics, comparable to that of the Vedic texts.
Consider for instance the case of the Pallava epigraphy, where a share of the land grant is predicated to the livelihood of a reader of the epic (‘vāśippavanukku’)121: we know that the epic was not “read” (√vāci [?], to read; not in DED), much less recited to an audience. To judge from the well-founded latter day praxis of the craft, a verse or a group of verses would be read or declaimed (rather than ‘recited’ with its Vedic connotation of proper accentuation and exact phonology) by the discourser to expatiate on issues of right and wrong, right conduct at moments of ethical or moral ambiguity, with, as we know, a good deal of sophistry and expostulation. A sample of such exposition is in fact a regular weekly column in the Hindu newspaper, appearing in the back page of the newspaper.122 We should contrast this with the example of the Homeric epics and public recitations of portions of the epics in the Pan-Atheniam festival in Athens. Plato’s Ion (530B2) makes it clear that the rhapsodes merely recited, if performatively, stretches of verses from the Homeric epics on stage in competition or contest with other rhapsodes123: no commentarial discourses followed the recitation. In the Indian example, we know that the praxis is completely different, the discourser reading from a written (printed, today) copy of the epic verse or passage from the epic as a take off strategy, as a point of departure, to pass on to his many homilies and casuistries on matters related and unrelated to the epic verses. In turn, we must contrast this with the tape recorder-like fidelity of recitations of the Vedic verses in Vedic rituals among the same people, in the same tradition. In other words, we may rule out oral tradition as a means in the transmission of the epic, both in time and space.
In addition, the parva-based transmission of the Mahābhārata text would have made the mastery of the entire epic to a memorial tradition impossible—the parva transmission itself being a consequence of the literate tradition, it should be added. One of the discoveries made by the Poona editors during the preparation of the Critical Edition was that the transmission of the epic was often along individual parvans, rather than the entire text of the epic, an inevitable condition with a text of the size of the Mahābhārata. It makes no sense to think that just one or two parvans would be mastered in oral tradition and transmitted as such. We could add parenthetically that if all parvans of the epic are found in a given resource center, then the text tradition of the center in question must be generally unimpeachable. This is what we find in the case of both the Pūrvaśikhā and Aparaśikhā Brahmans: each of these groups could have assembled a complete 24-parvan Southern Recension Mahābhārata text, as indeed they did. We have a complete verse-to-verse translation of the Pūrvaśikhā-Malayalam version of the Mahābhārata into Malayalam by the prince Kuññikkuṭṭi Tamburān in 1904-07; we have P.P.S. Sastri’s Kumbakonam edition of the Aparaśikhā Southern Recension in 1933, assembled from the Tamil (Grantha)-Telugu version of the Mahābhārata, from the Sarasvatī Mahāl Library: P.P.S.Sastri was the director of the library.124 For all these reasons, we can discount the possibility of an oral archivization and transmission of the Mahābhārata epic—both vertically in time from generation to generation and horizontally, across geographical space, from northern India to other parts. Indirectly, this supports the Hiltebeitel (2001: 20-21) thesis of a committee-based redaction of the entire corpus,125 a script driving, perhaps, the redactorial process. It is easy to see that the only script that offers itself is the Southern Brāhmī in Mahadevan’s chart, providing the conveyance of the *Śārada text to South India with the Pūrvaśikhā Brahmans, the latter group, with strong adherence to the Jaiminīya tradition and thus frame narratives, originally perhaps even part of the Hiltebeitel committee. Let us note that the sister script, Tamil Brāhmī, has already traveled southward independently with the Jain monks, who fashion this script by 3rd BCE to meet the demands of Tamil phonology, a point that cannot be overemphasized. That is, in effect, this script, the script of the Sangam poetry, cannot carry the full range of the sounds of the Sanskrit language and literature, ruling itself out for the transcription of the *Pūrvaśikhā SR Mahābhārata, although attested in the Tamil country by 3rd century BCE. The only script that possesses at the same time attestation in peninsular India, albeit late in Grantha script, by 5th century CE, and the ability to carry the full range of Sanskrit phonetics, is the *Southern Brāhmī script.
Once we accept this, many known and stray facts fall in place. The Pūrvaśikhā Brahmans depart the antarvedi area of the Gangā-Yamunā doab, with the *Śārada text in the Southern Brāhmī script, ca. 150 BCE. At and after their arrival in the Tamil country, they participate in the creation of the Sangam literature in the Tamil-Brāhmī script, already in use in the area, having been created earlier by the Jain monks. They also create the *Pūrvaśikhā text of the Mahābhārata in the Southern Brāhmī script, over the next several centuries. In other words, we must assume a sort of di-graphia,126 equivalent to diglossia, but in the realm of scripts, among the Pūrvaśikhā Brahmans, using the Tamil Brāhmī script for writing in Tamil and the Southern Brāhmī script to write in Sanskrit. 90 The rise of the Southern Recension text is proof positive for this: the text exists as a physical object, each of the 18 parvans of the Northern Recension worked over; material adapted from khilā (“appendix”) sections to re-fashion the main parts of the epic, as in the peroration of Bhīṣma on behalf of Kṛṣṇa in the Sabhāparvan; with several episodes transposed, the whole epic becoming more Brahmanical than the already Brahmanical Northern Recension and attaining a 24-parvan extent in its final form. It is not enough if we imagine the process in the abstract: we must account for the human agencies behind the process and the possible scripts that could meet the demands of a Sanskrit phonology. It is thus that the Southern Brāhmī script evolves into the Grantha script, over the half- millennium or so. The royal epigraphy of the three Tamil kingdoms in the area continues in the meanwhile to be in the Tamil-Brāhmī script, a practice already established by the Jain monks, with the “unique” adoption of a northern Brāhmī script for the non-Sanskrit, Dravidian phonology of Tamil, attested in a total of 70 inscriptions in the Pāṇtiyan kingdom, 17 in the Cēra kingdom, 5 in the Toṇṭai region, and 4 in the Cōḷa area, from 3rd BCE to 6th CE (Mahadevan 2003: 134). In the meanwhile, the Southern Brāhmī script, the script of the Mahābhārata epic, remains with the Pūrvaśikhā Brahmans, becoming the Grantha script in time and giving us the *Pūrvaśikhā text by the Kaḷabhra Interregnum. The date of the first appearance of the Grantha script in inscription supports this, the early 6th century CE, a century or so before the Old Kannada and Telugu scripts. Let us keep in mind the pakaḷiya attestations, and thus a Pūrvaśikhā presence, in the Tonṭaimanṭalam region during the Pallava period.
In other words, the Southern Brāhmī-Grantha script, say *Grantha script, is a paleographic counterpart of our *Pūrvaśikhā SR text. At the Kaḷabhra Interregnum, the future Nambudiri Pūrvaśikhās take both the *Pūrvaśikhā text of the Southern Recension and the *Grantha script to the Malabar area over the Palghat gaps, creating the Āryeḻuttu from the Grantha and the resident Vaṭṭeḻuttu scripts from 10th to 16th centuries. Indeed, the area of the Āryeḻuttu script shows itself clearly as an intrusive wake in the linguistic map of Kerala, formed by the arrival of the Nambudiri Pūrvaśikhās through the Palghat gaps, with Vaṭṭeḻuttu in use in areas both to the north, as Kōleḻuttu, a form of Vaṭṭeḻuttu, and Vaṭṭeḻuttu proper in the south, in the historical Travancore-Cochin region (Map IV). Both the *Pūrvaśikhā text, now distinctly as the Σ-text, and its Grantha script stay behind in the Tamil country, with the Śōḻiya Pūrvaśikhās. They create from the Grantha script of the epic and the Vaṭṭeḻuttu script of the Tamil Brāhmī family the extant historical Tamil script, the script of the Āḻvār (and Nāyanār) poetry.
We are now in the Pallava period of Tamil history and the arrival of the Aparaśikhā Brahmans, from 4th CE onward. There can be little doubt that the Aparaśikhā Brahmans were a literate group, allowing writing, unlike the Pūrvaśikhās, to enter even their Śrauta praxises. And the early Pallava epigraphy shows the script to have been the “Brāhmī Script of the Southern Class” (Mahalingam: 29-30).127 By the mature Pallava period, the Sanskrit parts of the Copper Plate paleography are in the Grantha script and Tamil parts, in historical Tamil script, the common script of the region, created from the Grantha and Vaṭṭeḻuttu scripts. As with the precedents of the Śrīvaiṣṇavism and the Southern Recension Mahābhārata, the Aparaśikhā Brahmans adapt themselves to the host traditions, in the matter of the writing systems as well. Is there a trace of the Aparaśikhā script that came with them, a counterpart to the σ-text in the final paleographical picture of the Tamil country? Sure enough: as William Bright notes (1998: 45) “[I]n the eighth century (CE) a competing script came into use for Tamil—probably reflecting a northern variety of Brāhmī, but with strong influence from the Grantha.” It needs to be scarcely added that the eighth century marks the arrival of the Aparaśikhā Brahmans in large numbers, with the rise of grāmadeya of 108 families, and we have our σ-script.
D. iv. The Brahmans, the Sanskrit Epics and Paleography
The famed Laurentianus codex of the plays of Sophocles,128 in the early “miniscule” style of writing, six plays and a fragment out of, it is thought, a total of 120 plays the playwright wrote through his long life in Athens, from 495-406 BCE, is dated to the 11th century CE. It was made in a Byzantium scriptorium from an eighth century CE archetype, with five extra lines on each page and enough marginal space for the scholia, already, it would seem, a set practice in the tradition of manuscript transmission of Western classical texts. It was acquired in Byzantium by Giovanni Aurispa, a Sicilian manuscript collector and dealer, between 1422 and 1423, and sent in advance of his own journey with an additional 238 volumes back to Florence, to Niccolo dé Niccoli, a prominent member of the group which surrounded Cossimo dé Medici in Florence. It lay in the Medici collections till 1523, traveling then to Rome with the Medici Pope, Clement VII, when he built the extant Florence Laurentian library to receive them. Another edition of the Sophocles plays appeared in the meanwhile, in 1502, in Venice, also from other Byzantium manuscripts, dating from 14th century CE, published by Aldo Munuzio, but in ignorance of and thus without consultation with the Laurentian manuscript. The Aldine text held sway till the second Juntine edition of 1547, the first Juntine edition having been published in 1522 largely based on the Aldine edition of 1502. The second Juntine edition of 1547 incorporates the codex Laurentianus of the Sophocles plays for the first time into the textual tradition the plays, thereby and thereafter making codex Laurentaianus the basis for the editio princeps of the Sophocles textual history. I provide this excursus into the textual history of the plays of Sophocles, not, as it might seem at first sight, to draw contrast between the histories of transmission of texts between east and west, the precision of the latter and the looseness or waywardness of the former but rather to show that an equally sagacious narrative of the transmission of texts is possible for the family of the Mahāhārata texts and manuscripts, if the right questions are posed and rational answers arrived at. Far too often, a regional text is taken for granted, given a disembodied existence, as if the epic unearthed itself there like the Copper Plate inscriptions, outside the realm of the questions that have governed this investigation. Both Brahman groups can be concretely linked to the textual history of the Southern Recension of the Mahābhārata epic. Moreover, we see that an adequate narrative of its formation can be obtained from the history of the paleography of the two major families of scripts of the region, the Sanskrit-able Southern Brāhmī script and the Tamil-able Tamil Brāhmī script—in other words, a literate version of texts, pointing to the fallacy of the idea of nebulousness, or worse, the absence of “texts”, in the east. The Ṛgveda all by itself is a constant and eternal repudiation of this fallacy, remaining an oral text for all practical purposes to this day among the Brahmans of this investigation. However, even Sukthankar echoes such a sentiment in his persistent invocation of the difficulty of the creation of a CE of the Mahābhārata with his reiteration, surely once too often, of the sui generis nature of the epic. There is no doubt the epic is sui generis, but it is so in the manner of most archaic texts.
This is the larger context in which I have framed the above argument that brings together three items in an algorithmic relationship, the Brahmans, the Sanskrit epics and their various scripts, the three irreducible correlates. It is quite true that we cannot conjure the Byzantine scriptoriums in the various points of interest in the textual history of the Mahābhārata—a point, ca.150 BCE, in the erstwhile realm of the Kuru-Pāñcāla chieftains and kings–Witzel’s Brahman kings promoting the Śrauta traditions–of the gathering of Hiltebeitel’s Brahman committee and the resulting *Śārada codex; a Sangam locale later, ca.100-400 CE, where the *Pūrvaśikhā Mahābhāratha was created; or a Nāyaka facility where the Aparaśikhā text took shape. What I have tried to show above is that only because some analogues of these facilities existed at these and other such relevant geographical points do we have the extant manuscripts of the different text- traditions of the Mahābhārata.
First of all, the analogues to the vellum parchments of the Byzantine scriptoriums. I have claimed above that the Pūrvaśikhā Brahmans left the antarvedi area of the Ganga- Yamuna area with the Mahābhārata epic, a version close to the Śārada text. What was the epic written on? I believe that we can rule out leather as the physical manuscript: *Śārada text was close to 75,000 verses and it is difficult to imagine enough leather for this much text. The būrjapatra is a choice for the material, and it appears as an item of trade in the Rāmāyaṇa.129 However, its supply, available only in birch forests 7500 feet high in the Kashmiri-Himalayan mountains may well be as rare as the Soma of the Mujāvat mountains. More likely, the physical manuscript would be the palm leaf linked to an ink-quill technology. Once it reaches peninsular India, the palm leaves can readily be imagined to take its place, and considering the sheer size of the text, it is even possible that the first transcription of the *Śārada codex in būrjapatra or palm leaf into the traditional peninsular palm leaves based on an iron stylus technology may well be the beginning of the process of the revision of the *Sārada text into what becomes the first ornate *Pūrvaśikhā text of the Southern Recension. We must keep in mind that by now, as Mahadevan notes,130 the Tamil society has become truly literate and the use of palm leaves for writing, pervasive—leading, indeed, as I note above, to the circular shape of the Tamil-Brāhmī script as it becomes Vaṭṭeḻuttu. Professional scribes, the equivalent of the personnel of the Byzantine scriptoriums, must have been widely available, extant in the 1950’s in my memory in Kerala as recorders–directly on palm leaves with iron stylus held in a closed right fist–of the horoscopes of new born babies, when pen and paper had become de rigueur in our other lives.131
Second, the script. If we accept that the Mahābhārata tradition is literate, then we have to deal with issues relating to a script in which the corpus was copied—in either būrjapatra in the north and palm leaf in peninsular India. An alternative, of course, is to imagine that the epic was in an oral tradition all the way to the dawn of the CE, as Fitzgerald intimated to me,132 close to 100,000 verses—without, however, a plausible infrastructure to support or maintain it in oral tradition. As already noted above, an institutionalized oral tradition was never part of the transmission of the epic, except perhaps at its origins. Things clarify themselves exemplarily once we cross this Rubicon. We see that, for the development of the Southern Recension in the physical medium of the palm leaf, the only relevant script is the Southern Brāhmī script. Its sister script, the Tamil Brāhmī script is already attested in the Tamil country by 3rd century BCE, its archetype having left northern India with the Jain monks some considerable time before—a century or so, as Mahadevan suggests (159)–for the Jain monks to develop from a Sanskrit-based writing system a script appropriate for Tamil phonology. However, the epic did not come to the Tamil country with the Jains, but with a group of Brahmans, almost two centuries later, by the dawn of the Common Era and the Sangam poetry, into an area already widely literate with the Tamil Brāhmī script. This is the logic—a Sanskrit text being made from one version to another—that forces us to accept the reality of the Southern Brāhmī as the script of the epic, and that it came with the Pūrvaśikhā Brahmans, the latter being the di-graphic human agency behind both the Sangam poems, in their Brahmanical contents and authorships, and the Southern Recension text of the Mahābhārata epic. We do not have an alternative explanation in the present state of our knowledge.
Third, the Brahmans: the analogues of Niccolo dé Niccoli and Cossimo dé Medici of the Sophocles text history. Both groups of Brahmans, Pūrvaśikhā as well as Aparaśikhā, were full equivalents to the Renaissance figures, in the matter of the transmission of the texts and literate scholarship. More than this, the really important point to note is that the infrastructure that served the transmission of the texts in South Asia was analogous, and of a high order. Both groups of Brahmans above brought the śrauta traditions of Vedism to the peninsular India, the first group, the Pūrvaśikhās by the beginnings of the Common Era and maintaining them still in a live oral tradition, and the second, Aparaśikhā Brahmans by the Pallava-Cōḷa periods, an entirely different tradition derived from a later corpus of Vedic texts and in a partly literate state. Indeed, this demands an infrastructure of far greater complexity than that needed to run the Byzantine scriptoriums. First and foremost, it needs a specific tri-Vedic axis of praxis: the hautram of a specific school of Ṛgvedic texts, the ādhvaryam likewise of a specific Yajurveda tradition and, third, easily the most important of the three, the audgātram of a specific Sāmaveda tradition—all institutionalized in the family-based svādhyāya system. Migrations of Brahman groups who have sustained a Śrauta tradition could only have been well-organized and systematized with the sort of sophisticated infrastructure such as the one we are led to imagine for Byzantium or Florence.
A large part of the infrastructure would be linked naturally to the demands and praxis of the Vedic tradition, the mastery of the three ritual Vedas in the first place and their immense and baroque viniyoga deployments in the rituals—demanding 16 priests for the śrauta ritual. We know that the śrauta ritual demands a rehearsal of some six months,133 as observed in its modern day performances. Even if we allow a shorter period for preparation and rehearsals from constant and regular practice, it would be nearly the occupation of an entire year. In other words, the two Brahman groups in question here, Pūrvaśikhā or Aparaśikhā, must be imagined as engaged in śrauta matters most of the year, performing the śrauta ritual every year at vernal equinox on their centuries-old migrations southward.134 The Assaka Soma ritual of the Suttanipada, possibly, is one such example. That they did so is proved by the survival of the śrauta Vedism in both groups, each distinct and autonomous. For instance, we know that the śrauta tradition with the Vādhūla school of the Yajurveda meeting the praxis of ādhvaryam has been extinct among the Nambudiri Pūrvaśikhās since the beginnings of 20th century CE 135: it also means that it had survived among them till then, from 5th century BCE, at the latest. To consider another example, a śrauta tradition is altogether no longer extant among the Tamil-speaking Śōḻiya Pūrvaśikhās,136 but we know from the Karandai Plates that it existed among them till 1029-30 CE, presumably in a live and continuous tradition from its origins.
To throw in the Mahābhārata epic into this infrastructure of transmission of systematized knowledge is to ask a small camel–okay, a large one–into the tent, albeit in a literate transcript in a generally oral tent. Once we accept the formation of the Mahābhārata in its present form and extent, and its canonical status as the fifth Veda, we cannot separate it from the Brahman groups of the type we encounter above. We must recall here that the founding myth of the Mahābhārata is a śrauta ritual, the Janamejaya Sarpa Sattra. This represents a Brahman possession of the epic, perhaps not wholly disconnected from the Pūrvaśikhā Brahmans of the above account in that Vyāsa, the master composer of the epic and a Parāśara Brahman, appears as part of the sadasya of the Śrauta ritual, an office unique to the Kauṣītaki hautram of the Pūrvaśikhā Śrauta axis and second, the hyper-developed frame narratives among the Jaiminīya groups, also part of the Pūrvaśikhā matrix: whether it was also at the same time a Bhṛgu usurpation of the epic is, I believe, not a wholly closed question.137 Brahman groups with the sort of learning infrastructure, or learning quotient, as above, would also keep the text in transmission, but as a literate transcript in an otherwise still predominantly oral culture. A literate artifact means a script, and we see that appropriate and relevant paleography is attested in both Brahman groups.
Lastly, we should resist the ease of imagination a disembodied regional version found in situ in isolated points of South Asia affords us, as in an abstract statement like “The Mahābhārata epic is found in its shortest Southern Recension in Kerala.” To subject such a statement to an Occam razor analysis, an analysis of its irreducible physical, areal correlates—the script, the physical form of the manuscript, the extent of the epic itself, the human agencies behind the texts—in terms, further, of their final filiations, is to arrive at the conclusions reached above: that the Mahābhārata, substantially the Śarada codex text of the CE, or the *Sarada text in my scheme, left the antarvedi area of northern South Asia ca.150 BCE with the Pūrvaśikhā Brahmans in a *Southern Brāhmī script in possibly the būrjapatra manuscript or the palm leaf manuscript of northern India, both using an ink-quill technology of writing;138 they created the *Pūrvaśikhā text in the Tamil-Kerala country from this in the half millennium after arrival, the recensional change from the *Sarada to *Pūrvaśikha probably taking place in the process of transcription from the northern manuscripts to the palm leaf manuscript of the South with the stylus technology, the original *Southern Brāhmī script becoming gradually the Grantha script in the process; a *Pūrvaśikhā text moves to the present territory of Malabar in Kerala at the Kaḷabhra Interregnum and comes to Poona for collation purposes toward the creation of the Poona CE; a *Pūrvaśikhā text remains in the Tamil country as the Σ-text to host Sukthankar’s σ-text, that is, playing host to the Aparaśikhā immigrants and to their Northern Recension text, creating eventually the Tamil (Grantha)-Telugu version of the Southern Recension.