Moving Targets? Texts, language, archaeology and history in the Late Vedic and early Buddhist periods

The Late Vedic and earliest Buddhist texts are investigated to indicate their relative historical layering. Besides the texts themselves, their language, place names, archaeological and their inherent historical background are brought to bear. These data and those on some historical contemporaries of the Buddha do not indicate a correlation with late Vedic personalities and texts. A certain period of time separates both corpora.

However, some of these questions may have been put the wrong way. As we will see, we rather are dealing here with two moving targets. Both text corpora have still not been explored sufficiently well as to establish either the beginning of actual Buddhist text formation, or the end of Vedic text formation and the closure of the Vedic corpus. 2 This has recently emphasized by Oskar von Hinüber: "… early Buddhist historiography … is deplorably absent…" and later on, "our sources never allow us to go beyond more or less likely or probable conclusions about the roots of the texts that reach far back into period of early Buddhism", adding the same caveat for the date of the closure of a text. 3 As a further caveat, it must be underlined that I leave aside, for the most part, the development of thought, philosphy and religion. It remains a difficult undertaking to trace their multiple strands, impossible to carry out in a brief paper. 4 Rather, I will concentrate on some aspects of archaeology, material culture, language, society and contemporary history. § 2 Materials In order to establish what occurred in the various parts of North India, and during which periods, both the locations of the individual texts and their relative or, better, their absolute dates need to be ascertained. Only on this basis it will be possible to illuminate what each text or group of texts from the various geographical regions and the various time levels involved can tell. However, in spite of some recent advances, much work still needs to be done in this respect.
The location of most Vedic texts of this period has been established fairly well over the past few decades. 5 However, it is still difficult to pinpoint the geographical location individual Pāli texts, even if one can be generally sure that the earliest texts of the canon were composed in the eastern part of Northern India, mainly in the Kosala (Audh), Kāśi, Bihar and Magadha areas. 6 The relative timeframe of the Vedic texts, however, is well established: they are layered in five levels 7 from the Ṛgveda down to the oldest Upaniṣads (Bṛhadāraṇyaka Upaniṣad, Jaiminīya Upaniṣad Brāhmaṇa, Chāndogya Upaniṣad). Yet, there still is no reliable absolute dating for these texts.
However, the language and content of Late Vedic texts, seen in the earliest Upaniṣads, overlap with those of late Brāhmaṇa texts, such as the first section of the (JB 1.1-1.65) or the independent appendix to the Taittirīya texts, the Vādhūla Anvākhyāna. They point to a living Vedic language at a time when the Buddha (like later on, Pāṇini) still could call Vedic Sanskrit "chandas" ('metrical'). 8 This is the period of the Late Vedic 'southeastern Koine.' 9 Both language and shared stories (both little researched) 10 point to the same period. The area is congruent with the Kosala, Kāśi, Vajji and Magadha territories of the early Buddhist texts.
There also are some small sections in the Late Vedic texts Upaniṣads that can be suspected to be post-Buddha, 11 and the final form of Vedic texts seems to have undergone some minor phonetic shifting and adjustment 12 15 To establish which were the earliest Buddhist texts and their layering still is a work in progress. However, the Pātimokkha formulas, to be recited monthly, or a large section of the Suttanipāta belong to this level. 16 21 It is also clear that later on a transposition of these texts in a «Buddhist Middle Indic» has taken place into the western 'literary' language, Pāli, 22 and that these texts were then collected and redacted at the so-called Third Council under the great King Aśoka, at Pāṭaliputra, c. 250 BCE. Some additional texts were included at least until, and sometimes beyond, the supposedly first written version emerged in Sri Lanka, shortly before the beginning of the common era 23 and more than a century before the extant manuscripts in another MIA language, Gāndhārī. 24 Some residues of the eastern dialect are still visible in our current Pāli texts, a problem that O. v. Hinüber has contributed to substantially. 25 However, the layering of the extant Pāli texts is not as well established as that of the Vedic canon. Again, O. v. Hinüber has made important contributions to this problem, but the large extent of the materials -many of them having multiple internal layering -does not yet allow for a comprehensive picture and summary, --though we have such important studies as that of the Suttanipāta. 26 In general we can rely on the first canonization of the Pāli texts in Aśoka's time around 250 BCE.
We must therefore extrapolate backwards to the life time of the Buddha, c. 460-380 BCE, in order to reach the earliest form of his sermons and speeches. This will also remain a problem in this paper throughout. I will thus clearly distinguish between «earliest Buddhist texts» of the Buddha's life time and those of the Pāli canon some 150 years later. 27 One way to get a closer handle on the problem is to investigate the language of both corpora closely, as language usually is a good yard stick for the age of a text, --that is, unless an author intentionally used archaic forms. 28 Unfortunately, this kind of evidence is again clear in the Veda, but less so in Pāli texts, as these have constantly been updated and, in part, even Sanskritized. 29 One has occasionally tried to achieve a historical layering of Pāli texts. 30 37 Other items such as the use of the perfect in narrative contexts stand out. In Late Vedic texts of the Center and East of northern India, the perfect replaced the earlier use of the imperfect. 38 This is standard in texts such as the ŚB and BĀU. However, the JB, largely composed and collected in the western lands south of the rivers Yamunā and Gaṅgā, had a mixed use of imperfect (for traditional, mythological tales) and perfect (for reports of more recent occurences) in natural speech, with a strong showing of the Aorist, especially when reporting facts with results in the present (the former domain of the perfect). 39 The latter situation can be seen as the predecessor to the complete domination of the aorist in reporting the past in the Pāli, which is, after all, like the JB's, a western MIA language. The contrast between the few (5) 47 Nagarin 'one who has/is characterized by a nagara' occurs only in a name, and always as the same person, Nagarin Jānaśruteya. He appears in Jaiminīya and Aitareya Brāhmaṇa as well as in Jaiminīya Upaniṣad Brāhmaṇa, 48 as belonging to the Dālbhya clan, once along with with Āruṇi, 49 The absence of cities is also typical for the earliest level of Buddhist texts: the Buddha stayed and taught in villages, not cities. 50 Notable are the place names in early Buddhist texts such as Rāja-gaha ('king's house'), or Pāṭali-gāma, the modern Patna. 51 2, 6.4.1sqq,  6.4.3, 6.5.4, 6.6.1sqq, 6.14.2, Indikā 4.10, 19 AAWG 189, 194, 222, Göttingen 1991, 1992, 1997 However, to begin with, the designation ajātaśatru «one who has 'unborn' (non-existent) enemies» is not rare and is indeed found since the Ṛgveda, 89 It can have been the name of many men ever since. One of them, the Kuru King Ajātaśatru, is definitely to be excluded as being identical with Ajātasattu. He is the son of a Medhātithi (a maidhātitha), not of Bimbisāra or whathever this Magadha king may have been called in Vedic. 90 Yet, the Veda also has one mentioning of a son of a certain Ajātraśatru, called Bhadrasena Ājātaśatrava. 91 This occurs in a ritual discussion of the Sautrāmaṇi ritual. He was bewitched by one Āruṇi (whose son Kusurbinda has an autochthonous name). The famous Yājñavalkya is said to have responded to this kind of sorcery. This would place the tale in pre-Magadhan times.
It is however notable that Ajātaśatru is not called a king here. The appearance of Āruṇi could point to the Brāhmaṇa/Upaniṣad period 92 but some Āruṇis are also mentioned in an early Saṃhitā text. 93 In addition, the ŚB passage ocurs in the closing statements of the discussion of the Sautrāmaṇi, just before the dakṣiṇās (ŚB 5.5.5. [16][17][18], and thus can have been added any time before the (late) redaction of the text. 94 Based on these data, Bhadrasena can be the son of any of the Ajātaśatrus in the Veda, mentioned or not, since the RV, though the juxtaposition of Yājñavalkya weighs in favor of a late Vedic timeframe of the tale. 95 Importantly, the late Aitareya-Brāhmaṇa already speaks of kings not belonging to the Kṣatriya class -which is typical for the Nanda and Maurya dynasties and will have been so for the earlier Magadha kings as well. 89 In sum, both the names Ajātaśatru   Cumulative evidence --historical, archaeological and textual--of the Late Vedic and early Buddhist texts therefore points to a clear time gap between both text corpora and the time periods they depict.
In sum, the results of this limited investigation, which intentionally excluded the development of thought, uphold the "traditional" view of several consecutive linguistic, textual and historical layers from Vedic to the earliest Buddhist texts. We can be fairly certain, that this sums up "wie es wirklich gewesen" --evam etaṃ bhūtapubbam. 98