Houben 2016 -- From Fuzzy-Edged "Family-Veda" to the Canonical Ś¯ akhas of the Catur-Veda: Structures and Tangible Traces

-DQ(0+RXEHQ ³)URP)X]]\(GJHG³)DPLO\9HGD´WRWKH&DQRQLFDOĝƗNKƗVRIWKH&DWXU9HGD 6WUXFWXUHVDQG7DQJLEOH7UDFHV´ $SSHDUHGLQ9HGLFĝƗNKƗVSDVWSUHVHQWIXWXUH± 3URFHHGLQJVRIWKH)LIWK,QWHUQDWLRQDO9HGLF:RUNVKRS%XFKDUHVW (GLWHGE\-DQ(0+RXEHQ-XOLHWD5RWDUX0LFKDHO:LW]HO +262SHUD0LQRUD,;±&DPEULGJH0DVVSS From Fuzzy-Edged “Family-Veda” to the Canonical Śākhas of the Catur-Veda: Structures and Tangible Traces Jan E.M. Houben In memory of Prof. J.C. Heesterman (����-����) �. Veda and Vedic schools: a remarkable phenomenon �.�. The formation of the Veda and the development of Vedic schools in ancient India constitute a complex phenomenon that is, from a global perspective, entirely unique in character and extent, even if components of this complex phenomenon can play important roles in the scientific study of what has recently been termed “natural experiments of his� tory.”1 In fact, such components have already played a role in comparative anthropological studies, unfortunately often on the basis of incomplete or even incorrect information. The formation of Vedic schools, for instance, figures among the topics addressed by Randall Collins in his Sociology of Philosophies (����, ���-���), in the chapter devoted to India, one of “the world’s three great indigenous intellectual traditions” (����, ���). Collins’ conclusion of this brief section may be appropriate: “Strong positions divide, weak posi� tions unite” (����, ���). In other words: an intellectual tradition tends to subdivide into branches when it is strong and favoured by prevailing conditions; but when it is weak and subject to unfavourable conditions, we see syncretism and unification of divergent and formerly strongly competitive schools. However, the misrepresentations in Collins’ brief account on which he bases his conclusion – which, not unexpectedly, fits his general theory that should apply to all intellectual traditions – are so numerous that their enumeration cannot be undertaken here.2 1 Jared Diamond’s and James A. Robinson’s important methodological observation deserves to be briefly quoted here: “The controlled and replicated laboratory experiment in which the experimenter directly manipulates variables, is often considered the hallmark of the scientific method. . . . That fact misleads laboratory scientists into looking down on fields of science that cannot employ manipulative experiments. But the cruel realtiy is that manipulative experiments are impossible in many fields widely admitted to be sciences. That impossibility holds for any science concerned with the past, such as evolutionary biology, paleontology, epidemiology, historical geology, and astronomy; one cannot manipulate the past. . . . A technique that frequently proves fruitful in these historical disciplines is the so-called natural experiment or the comparative method.” (Diamond & Robinson ����, �-�) 2 For this brief section, Collins has apparently relied heavily on secondary sources which are themselves based not directly on primary sources but on translations (e.g. Stutley ����). It is Jan E.M. Houben Another component in the formation of the Veda and the development of Vedic schools is the transmission of knowledge, generally admitted by specialists to have been entirely oral in early times, later on also in written form. The case of the transmission of the Veda drew the attention of anthropologist J. Goody, as it did not match with his general theory primarily based on observations on the transmission of knowledge in some African tribes. On the basis of certain observations in secondary sources on the Vedas he finally succeeded in representing this case too as supporting his general theory.3 A reliable account of “the formation of the Veda and the development of Vedic schools in ancient India” is evidently a primary requirement, which precedes the probably important role which components in this phenomenon will be able to play in future studies of “natural experiments of history.” Contributing to such reliable account is the main purpose of the present study, but the perspective of possible comparative investigations will not be entirely neglected. From the point of view of Indian and Vedic studies, the importance of the problem of the formation of the Vedic schools or śākhās or branches of Vedic learning and ritual practice was expressed by Louis Renou as follows in his foundational work Les Ecoles Vediques et la Formation du Veda (����, ���f4 ): The problem of the śākhā is central to the problems of Vedic studies, and it is clear that if we would succeed in establishing on solid foundations the description and the lines of descent and interrelatedness of schools, we would at the same time know how the entire Vedism developed. Renou is quick to add an observation which will not come as a surprise to those who have a little more than just a superficial philological familiarity with Vedic literature (����, ���5 ): The enduring and unchanging character of the Vedic tradition is a myth and it is at the most valid for those few Saṁhitās which are equipped with a padapāt.ha and are preserved by a continuous and multiform recitation. Renou was, in fact, quite sceptical on the possibilities to really “succeed in establishing on solid foundations [my emphasis] the description and the lines of descent and interrelatedness of Vedic schools” and stated that “to hope to achieve this goal, however, is futile” (“espérer atteindre ce but, d’ailleurs, est vain”). Why was Renou so sceptical? not clear from where Collins drew that “the proprietors of the Atharvaveda” were “a coalition of magicians from the indigenous non-Aryan population” (����, ���). Collins may be right that “A reversal sets in” after about ��� B.C.E., which ends a period of “splits and rivalries among the Vedic schools” (ibid.). However, it is not clear on what basis he thinks that from that time onward “all the sects” are “merging into a common front of Vedic education” (ibid.). 3 Goody ����; ����, ���-���; discussions by Falk ����; ����, ���, and in Houben & Rath ����. 4 Renou ����, ���f: “Le problème de la śākhā est au centre des problèmes védiques, et il est clair que si l’on réussissait à établir sur des bases solides la description et la filiation des écoles, on saurait du même coup comment s’est développé l’ensemble du védisme.” Some of the gravest misunderstandings in the relevant sections and studies of Collins and Goody could have been prevented had they taken Renou’s work as starting point for their reflections. 5 Renou ����, ���: “La constance de la tradition védique est un mythe; elle vaut tout au plus pour les quelques Samhitā munies d’un padapāt.ha et préservées par une récitation continue et multiforme.” 160  From Fuzzy-Edged “Family-Veda” What remains of the literature, as vast as it is, does not capture the totality of the facts. To reason with those empty names which, in long, barely articulated enumerations, are given in texts that are more or less certain and of a relatively recent date, would amount to restoring a fictional history. And to ignore the tradition, retaining as acceptable only what actually remains, would amount to condemning oneself to another form of fiction.6 Another set of difficulties mentioned by Renou is “that most of the texts were open to additions, to modifications; that corruptions have spared so to speak none of them.”7 This problem had already been formulated by A.C. Burnell in ���� (p. xv) as follows: The Vedic literature as we possess it is unfortunately far from perfect; we have often only the later recasts of old works, or we have several of these, though but fragments of the works they are based on. �.� . In spite of Renou’s scepticism, Prof. Michael Witzel has proposed, four decades later, a strategy and a methodology to get a better grip on the authors and transmitters of Vedic texts and of the rituals these texts describe or presuppose. This would contribute to a better knowledge of the exponents of Vedic culture and to a better geographic and chronological “localization” of the Vedic schools (esp. Witzel ����, ����, ����a, ����b, ����; his recent study on “Gandhāra and the formation of the Vedic and Zoroastrian canons”, ����, contributes to the same line of research). The main parameters in the grids which Witzel proposes to set up for the Rg-veda (esp. in Witzel ����a and ����b), and, by extension, for all Vedic texts, concern, ˚ on the one hand, linguistic and textual structures and regularities, and, on the other hand, textual references to rivers and mountains and references to chiefs and poets who are often mutually linked by family relations representing a limited number of generations. From the “collapse” (or at least: transformation and disappearance) of the Indus civilization at around ���� BCE and the beginning of the use of iron (���� BCE), Witzel derives a broad chronology which some scholars, esp. those who prefer to perceive a greater role for ancient India, have challenged (see some of the discussions recorded in Bryant & Patton ����). From the references to rivers and mountains, Witzel derives a rough geography centered around the area of the “Greater Punjab” (currently in Pakistan and northwestern India). This rough geography is well-founded and it is accepted by a wide range of scholars, including those who contest Witzel’s broad chronology. The references to generations of chiefs and poets are probably correctly interpreted as references to extra-textual reality, but for these no independent dating is available. Witzel’s grid therefore mainly consists of a cluster of relative chronological relations and estimates, which as a whole has still insufficient anchorage in independently datable realities. Nevertheless, the methodology proposed by Witzel is basically sound as it leads to verifiable or falsifiable statements regarding the Vedic schools and their texts and rituals. 6 Renou ����, ���: “[C]e qui nous reste de la littérature, si vaste soit-il, ne permet pas de saisir la totalité des faits. Raisonner avec ces noms vides qu’en longues énumérations, à peine articulées, donnent des textes plus ou moins sûrs et de date relativement récente, c’est restaurer une histoire fictive. Et ignorer la tradition, ne gardant comme acquis que ce qui subsiste en fait, c’est se condamner a une autre forme de fiction.” 7 Renou ����, ���: “Un autre est que la plupart des textes ont été ouverts aux additions, aux remaniements; que les corruptions n’ont épargné pour ainsi dire aucun d’entre eux.” 161 Jan E.M. Houben As already suggested at the end of my study on “Vedic ritual as medium” (Houben ����), I therefore propose to expand and supplement Witzel’s grids by two additional parameters – “knowledge transmission” and “ritual” – that are related, on the one hand, to the existing web of relative chronological relations based on observable textual structures; and, on the other hand, to tangible traces that link, or, where further archeological or other research is required, promise to link, the grid of relationships to absolute chronology. �.�. Renou’s scepticism was legitimate, as it was based on the most thorough review of evidence and research available in his time. If we, at present, want to go beyond this scepticism, and try to follow the basically sound methodology which Witzel first proposed ca. �� years ago, we will have to confront, sooner or later, the question: what will count as establishing a description of the development of Vedic schools or some part of that development “on solid foundations”? I propose that what will count as “on solid foundations” should be based on: I. direct textual evidence (pratyaks.a) and inferences (anumāna) based on texts in their (either immediately attestable or, where needed and possible, reconstituted) context. II. cultural-historical comparison (anvaya-vyatireka). This will basically be a matter of inference (anumāna) that takes into account close cultural parallels. Next, we have to be realistic and be aware that in historical reconstruction some extrapolation from better known situations and filling in of lacuna in our information will be unavoidable, implicitly or explicitly. If this is so, it is better to try to be explicit about it. Hence: III. If there is no textual or comparative basis that can enable us to fulfil these conditions, remaining gaps in our reconstructions may be filled up through estimates and hypothetical scenarios, but these should never contradict I and preferably neither II. Basically, this too will be a matter of inference (anumāna), but one for which no direct use can be made of close cultural parallels because of the uniqueness of the subject under consideration. �.�. Two further limitations I impose on the current exercise: First there is a spatial limitation: I will be focusing on what is happening within the Indian subcontinent in a wide sense of the word, so from what is at present North-West India, Pakistan, Afghanistan and the eastern-most fringes of Iran in the West to what is now West-Bengal, Bangla-Desh and Assam in the East; and from Kaśmir and the Himalaya in the North to Kanyakumārı̄ in the South. Second, a limitation in time is to be respected. It would not be wise to impose hard and fast chronological limits before hand, because these limits are among the things to be determined and about which there is still controversy. However, a limitation in time can be achieved if we take as starting point a period for which sufficient data are available and go from there gradually back in time till our search stops on account of one of two reasons: (a) we have found a chronological range where conditions and variables obtain that sufficiently explain “the lines of descent and interrelatedness of Vedic schools”; (b) although we found no satisfactory explanation for “the lines of descent and interrelatedness of Vedic schools” it has become clear that there is no chance to find any better explanation by going further back in time. On account of a general ideological pressure to find earlier origins for traditional Indian 162  From Fuzzy-Edged “Family-Veda” texts and institutions8 , the proposed method for a limit in time requires clarification. In our quest to understand the formation of the Veda and the Vedic śākhās, to understand the later periods is somehow a manageable problem as the available data are relatively abundant, even if they have not yet been sufficiently studied. However, it is a major challenge to understand the earlier periods which are naturally foundational for the later ones but for which solid evidence is most scarce. Our starting point must therefore be in a time when sufficient uncontrovertible evidence is available to make confident conclusions about the state of the matter. In practice, that starting point is the time of inscriptions touching on Vedic schools and Vedic teaching in the first millennium CE and the first half of the second millennium CE (cf. Scharfe ���� and Rath ����), when in addition manuscripts of central Vedic texts become available. Manuscripts of marginal Vedic texts are available from an earlier time, but their value is different from that of central Vedic texts. We therefore start from the second half of the first and the beginning of the second millennium CE, and go gradually back in time to the clear but relatively scanty evidence of Aśoka, third century BCE, where we find the earliest inscriptional references to baṁhan.as or brāhman.as; to the mediated evidence of the Buddha through texts that are securely established only a few centuries after him; and to the detailed evidence of Pān.ini, who can be dated only on the basis of indirect evidence, but at present nevertheless with considerable confidence in around the middle of the fourth century BCE.9 Working from there again further backward in time will require tight argumentation on the basis of the structures of the available texts and a continued search for tangible evidence, till we can define a space and time where conditions pertain that in all major respects match those apparently belonging to the creators and earliest transmitters (and schools) of Vedic texts. The evidence of later periods, for instance the time when inscriptions are available, does give some indication about earlier times when the same type of evidence – in this case inscriptional evidence – was not available. However, we have to find out whether some conversions are to be made. For instance, the mass of inscriptional evidence points to land grants often to Brahmins of specified śākhās and gotras. Inscriptional evidence therefore points to a time when the limiting factor for economical independence and stability needed for a life devoted to Vedic learning, teaching and ritual practice consisted of the possession of a smaller or bigger piece of land. However, from the textual evidence of early Vedic texts up to the Upanis.ads we know that the possession of land was at that time not the main problem, apparently because it was rather abundantly available. Instead, at that time the limiting factor for economical independence and stability needed for a life devoted to Vedic learning, teaching and ritual practice was the possession of cows. Yājñavalkya, at some undetermined date in the past, 8 As Randall Collins observed with regard to the history of Indian philosophical thought, there was a time when earlier competitions for better or more impressive doctrines (reflected in early Upanis.ads and in early Buddhism) give way to attempts to outshine the opponents by claims for more ancient cultures (����, ���): “Hindu and Buddhist texts now began to make extravagant claims for the antiquity of their cultures, the Buddhists by inventing cosmic incarnations of the Buddha who lived in prior eons . . . Now sets in the contest of ‘more ancient than thou’, which displaces the prestige of doctrinal innovations found among Upanishadic sages and in early Buddhism . . . ” On the tendency to date texts and institutions at earlier dates see also Bronkhorst ����. 9 Hinüber ����, ��; Falk ����, ���. 163 Jan E.M. Houben was happy to receive from king Janaka a gift of a thousand cows, but Brahmins of a later age were happy to receive a piece of land where they can keep their cow or cows and do some agriculture. Similarly, on the basis of the earliest ritual rules concerning the giving of the sacrificial honorarium or daks.in.ā we have to infer that the limiting factor for economical independence and stability needed for a life devoted to Vedic learning, teaching and ritual practice was, again, the possession of cows: this inference therefore matches the one made on the basis of the Upanis.adic story of king Janaka and Yājñavalkya. However, in one of the Pariśis.t.as of the Śuklayajurveda, the Mūlyādhyāya-pariśis.t.a studied by Frederick Smith in an article published in ����, we have a “conversion table” which explains, among other things, that one cow mentioned in ancient ritual rules can be exchanged by one gold coin in the time when this pariśis.t.a was composed. If we want to go back from later times to earlier times we have to make the same conversion in reverse, and draw our conclusions regarding the economic and ecological conditions of the earlier period. �.�. It should be added here that many divergent accounts have been proposed for the first beginnings of the development of the Veda: the earlier we go the more divergent are the proposed accounts of different scholars. Several alternative scenarios have been discussed in the already mentioned book edited by Edwin F. Bryant and Laurie L. Patton, The Indo-Aryan Controversy (����). Because we start here from later periods where evidence is relatively extensive and go back in time according to well-argued steps, we are able to steer clear of most of these discussions and their ideological overtones, discussions which become more vehement for the periods where less and less evidence is available. As we will see, our new parameters receive ample substantiation if we go back in time from Pān.ini and the Buddha till the first quarter of the second millennium B.C.E. This justifies, for our present purpose, an attitude of agnosticism regarding the representatives of Vedic culture or of predecessor(s) of Vedic culture from before that period. Alternative scenarios proposed in the mentioned discussions are therefore for us at present of no direct relevance except to the extent they are based on secured evidence and evidence-based inferences. �. The Large Outlines �.�. With ideological trench wars being fought regarding the oldest periods because of the symbolic capital at stake, and with numerous details remaining open for further determination not only in the oldest periods but also in the later ones, it is easily overlooked that there is, apart from various disagreements, a considerable area where most specialists are entirely in agreement. Except for the problem of the earliest beginnings, the large outlines and the internal, relative chronology of the develop- ments to be studied are relatively clear and rather uncontroversial. These are relatively clear and rather uncontroversial, from the beginnings of modern Vedic philology onwards and without much change in the subsequent one and a half century, in the work of pioneering specialists. Problems have come up, in those early days, and over the last hundred and fifty years, the very moment one tries to provide anchorage to the grid of relative chronological relations of the numerous Vedic texts, that is, when we want to link those relative chronological relations with external or absolute chronology. It is therefore important to be well aware that from the structure of the currently still available texts certain remarkable and significant, relative chronological relations 164  From Fuzzy-Edged “Family-Veda” are immediately clear: Brāhman.as presuppose and comment upon Saṁhitās and must be later than these and are in general indeed linguistically demonstrably later than these, as is evident in the Rgvedic tradition and in various schools of the Sāmaveda; a ˚ Śrautasūtra typically presupposes both the Saṁhitā and Brāhman.a of its own branch and to a smaller or greater extent also those of other Vedas and of other branches. Its relative lateness is confirmed by the linguistic forms it uses. In only a few cases doubts persist and on account of historical circumstances a text which according to structural parallels is expected to be later may turn out to be earlier in the light of other criteria. This applies to the relationship between the Brāhman.a, Śrautasūtra and Grhyasūtra, which are usually established in this historical and textual sequence in all Vedic ˚ schools. In the Atharvaveda, however, the sequence is different, with the Kauśikasūtra, which is similar to a Grhyasūtra, apparently preceding at least the Śrautasūtra (Caland ����, ����) or also ˚the Brāhman.a (Bloomfield ����). Which of these propositions should be accepted, or whether they should be reformulated, or entirely rejected is a different matter which need not be discussed here. Occasional exceptions to the easily observable regularities are there, but they are often explainable and incidental and as such only confirm the regularities. Whenever the addition of some new element or parameter to the already existing network of items and relations is investigated, there is a risk to get carried away by problematic details in connection with that element or parameter. It may, finally, emerge that the new element or parameter suits well, or, on the contrary, that it is to be rejected. Irrespective of the outcome of the investigation, it will in the first stage of the investigation be useful to remind oneself again of the large outlines that are already clear and relatively well established. �.�. With regard to the formation of the Veda and the development of Vedic schools, what can be regarded as well-established? In order to arrive at an overview we combine two groups of variables: (A) organizational structure of the texts: first RV-family books; next RV plus other Vedas; Vedic schools or Śākhas; ˚ ˚ (B) geographical area. Many crucial observations which add numerous details have been given by Witzel in the publications referred to. However, for our initial grid we focus on the general outlines while taking into account these two major and rather non-controversial parameters. Diagram I: The Large Outlines Are Clear (�) RV, around ���� BCE (estimate): Vedic literature and Vedic ritual are at first orga� ˚ nized according to expansive RV-families and clans: each parallel and competitive ˚ group has a similar array of textual materials; / Area: NW of Indian subcontinent: larger Punjab, (�) Late+Post-RV, after ���� BCE (estimate): Development of Yajurveda and of Śrauta-ritual˚ in which three, later four Vedas have a role: functional diversification and re-organization: some families focus (continue their focus) on Rg-recitation, others (acc. to gotra usually not different from RV-families) specialize ˚in Yajurvedic ritual knowledge, in Sāma-chanting, etc.; ˚ 165 Jan E.M. Houben / Area: Kuruks.etra and environment, (�) Late+Post-YV, from ��� BCE (estimate): Formation and development of specialized and, per Veda, mutually only marginally differring schools; / Area: first the north, next the whole subcontinent. �. The Large Outlines Plus Knowledge Transmission �.�. To these large outlines based on parameters A and B, we first add the new parameter (C) “Knowledge Transmission” some aspects of which in connection with Vedic ritual have been discussed earlier (Houben ����, ����). In the last few decades, modern scholarship has established the profound and multifaceted impact of the mode of “Knowledge Trans� mission” on intellectual traditions, on the arts and sciences, and on the construction of social and political communities. More particularly, the focus has been on print culture. Moreover, various studies have brought out that in an oral environment where no writing is used, traditions of knowledge and culture have distinctive features that change significantly once these oral traditions become literary traditions. In both groups of studies the focus has been on western and more specifically European history. The domain of manuscript culture and epigraphical writing, which lies in between the domains of orality and print culture, has received only marginal attention in contrast with the two other domains. The same applies to Asia in general and India in particular. For the Indian world contours and parameters of manuscript culture have now been explored in Houben & Rath ����. As for our understanding of print culture and its impact, Elizabeth Eisenstein remarked in her classical study The Printing Press as an Agent of Change (����, � note ��) that she uses the ‘print culture’ “to refer only to post-Gutenberg developments in the West. How printing affected pre-Gutenberg Asia must be left to others to investigate.” The importance of Eisenstein in several decades of study of print culture and its general focus on western and European history is clear from studies such as A. Johns (����) and the article N. Hudson ����. A pioneer on the implications of oral culture, in contrast to emerging forms of literacy, was Eric A. Havelock (����-����; publications ����, ����, ����), who established his insights on the basis of Greek antiquity. Walter J. Ong (����-����; publ. ����) and Jack Goody have added other areas (esp. Africa in the case of Goody) into the investigation of orality vis-à-vis literacy. What emerges from these studies is that the mode of knowledge transmission has multifaceted implications which are visible, inter alia, from the structure of a text: an orally composed and transmitted text is different from a text composed and transmitted within a culture of writing, and this is again different from a text composed and transmitted within print culture. The earliest written texts, for instance the dialogues of Plato, may still exhibit important characteristics of orality, which in the course of time give way to textual characteristics of written texts, later on to those of printed texts. What is here important is (�) these modes of knowledge transmission are associated with tangible forms of the text in the stage of manuscript and epigraphical writing and in the stage of print culture (unfortunately, these stages start relatively late with regard to Vedic texts in general and central Vedic texts, the Saṁhitās, in particular); (�) the complex of texts show an internal sequence, from apparently entirely oral (Saṁhitās) till later transitional phases and up to the time when writing is a fully accepted means to transmit, for instance, commentaries and manuals (paddhati s). We will here not be concerned with the stage when printing becomes significant for the transmission of Vedic texts in India, but with 166  From Fuzzy-Edged “Family-Veda” earlier stages. Following this internal sequence that can be deduced, approximately, from the general character of the texts, it is possible to reach a much earlier period than the earliest available written sources. However, it does not seem to be possible to provide any secure external dating for this oral phase. �.�. So far, the general insights based on the study of European and later the study of African conditions seem very well applicable in India too, and our main task would seem to be the application and elaboration of the general theories to India. However, the case of India, and especially the case of the transmission of the Veda, turns out to be quite unlike that of other traditions. As in several other traditions, the written tradition of the (central texts of the) Vedas, is preceded by an oral tradition of these texts. According to a thesis formulated in ���� by Jack Goody and Ian Watt, “science and even rationality originate from literacy.”10 In response to this thesis, generalized on the basis of a detailed study of, inevitably, only a limited number of concrete cases, Frits Staal (����) argued that it does not match the Vedic oral tradition in India and its role in the emergence of grammar and other sciences or knowledge systems. Staal distinguished “two traditions of transmission” in India: in one of these two the epics are at first orally transmitted and show then a certain “instability and change, at least on that verbatim level that is significantly called ‘literal’,” before they are written down and achieve “much greater stability” (����, ��). The other tradition of transmission concerns the Rgveda and a few other Vedic texts. The Rgveda, too, “may have gone ˚ of change that is not entirely dissimilar to˚ the oral vagaries of the epics: through a period but if it ever took place, it happened before its codification around ���� B.C.E. After that it was not written down, but entered the period of oral transmission . . . which fixed it for more than ���� years in the course of which it preserved as stable and solid a state as any textual ‘archetype’ ever has.” Apart from the dates which are only rough estimates, Staal’s position matches the one defended by Harry Falk (����; ����, ���) on the basis of other arguments. Referring to BaudhDhS �.�.�-�, Falk (����, ���) demonstrated that Vedic ritual culture gives unambiguous indications of an originally exclusively oral transmission and use of Vedic texts. Falk showed, moreover, that the form of the script (Brāhmı̄) as attested in the first Indian inscriptions (Aśoka’s pillar and rock edicts, mid �rd century BCE), was suitable to write Prakritic languages but not Sanskrit, and even less Vedic. This condition persisted for several centuries, till the script was gradually adapted to write Sanskrit (with characters such as r, consonant clusters). The slowness ˚ of the adaptation of the Brāhmı̄ script to write Sanskrit and the initial instability of attempts to solve problems in writing specific for Sanskrit thus indicate the absence even of an informal tradition of Sanskrit writing (hence also the absence of Sanskrit writing on perishable material), next to the writing of Prakrit which is epigraphically attested from the time of Aśoka. Vedic texts must therefore have been transmitted orally over long stretches of time. The special technique of recitation of the same text in continuous form (saṁhitā-pāt.ha), in 10 Goody & Watt ����, ���-���; the thesis is elaborated in subsequent work of Goody; for this brief formulation: Staal ����, ��-��; other refutations of the application of Goody & Watt’s thesis to India were formulted by Falk ����; ����, ��� and by Scharfe ����, ��: “Goody’s denial of the orality of much of the Indian tradition contradicts evidence from a multitude of sources”; Bronkhorst ����, however, does not agree with Staal and Falk and tends to accept the applicability of Goody’s ideas to India; he does not respond to the new arguments of Scharfe published in the same year; further discussion in Houben & Rath ����. 167 Jan E.M. Houben word-by-word form (padapāt.ha), and in the intermediate step-by-step form (krama-pāt.ha: two words are joined in continuous fashion, two by two, AB-BC-CD. . . ) has here con� tributed greatly to the exactness of the transmission. This technique, to which we will briefly refer as pada-plus-saṁhitā recitation, was already well known to the grammarian Pān.ini and his major commentator Patañjali, and is still adhered to in modern traditional schools for the study of Vedic texts. �.�. In order to illustrate the role the padapāt.ha and its recitation next to that of the saṁhitā-pāt.ha have played in the transmission of Vedic texts, Scharfe compared it with an entirely different mnemotechnique used in Greek and Roman Antiquity. The Greeks and Romans had developed a technique called “mnemonics”, by which they linked elements of an oration to features of a house or a street as one would encounter them on strolling through. Once considered a powerful and dangerous tool, its efficacy was later questioned: while it may help to remember the sequence of elements of an oration, it could hardly help one to remember the exact words or contents, and it fell into disuse, was later revived in the Middle Ages and finally abandoned for good in the Renaissance. Whatever efficiency this technique had—and its value cannot perhaps be denied altogether—may lie less in the vivid images employed than in the implied intensive occupation with the text: a manipulation of the text by which it is seen from different angles, so that it is engraved, as it were, deeper in the mind. That is exactly what the Pada-patha and the other modifications achieve. It is not that the reciter tries to reconstruct the Samhita-patha from the Pada-patha, but he remembers it better, because he playfully manipulated it. (Scharfe ����, ���) If Scharfe’s observation is correct, the intense occupation with the continuous text and the padapāt.ha would be the major reason for the resulting preciseness of the tradition. It would then not be necessary to insist on an almost absolute preciseness of the pada- and kramapāt.ha as Staal (cited above) did. It is rather the very practice of pada- and krama-recitation that would contribute to a better transmission of the text. That pada- and krama-recitation were developed quite late in the Śaunakı̄ya AV and that there was scope for variation and re-constitution has been demonstrated in detail by Deshpande (����). The padapāt.ha of the Rgveda seems to be rather stable and correct, but is not ˚ free from occasional wrong or problematic forms either (Oldenberg ����; Bronkhorst ����). It is important to note that the authors of the oldest Vedic texts, the hymns of the Rgveda, do not show the slightest interest in the word or pada as linguistic unit.11 It is ˚ difficult to say whether or not the Vedic poets were at all aware of the word as linguistic unit, but they show no awareness of it. They do have a keen interest in all levels of metrical units, from the smallest, aks.ára or syllable, to larger ones such as the arká ‘hymn’, vāká 11 There is, indeed, no dearth of attempts to trace the later concept of the pada ‘word’, analyzed and employed with so much sophistication in the grammatical tradition which we can follow from Pān.ini onwards (and which we also found in the Nirukta and in the Prātiśākhyas), back to the Veda, preferably the Rgveda. A detailed discussion of several such attempts is given in Thompson ����. The parallelism˚ of a hidden padá and hidden nā ´man in certain circumstances does not bestow the meaning ‘word’ to padá. 168  From Fuzzy-Edged “Family-Veda” ‘unit of recitation’. The word which in later Sanskrit will mean ‘word’ does occur in the Rgveda as padá, but there, if it has anything to do with language (i.e., if its meaning is ˚not within the range of ‘footstep’, ‘track’, ‘footing’, ‘abode’), it means the line or quarter of a verse.12 The association with nā´ma ‘name’ in verses such as RV �.��.� still keeps us at a considerable distance from a padá with the meaning ‘word’13 ˚ as it is found in later Sanskrit. Not surprisingly, early Vedic literature is also entirely ignorant of the padapāt.ha. A well-known author of a padapāt.ha, that of the Rgveda, is Śākalya, who is already mentioned in Pān.ini’s grammar as an earlier grammarian ˚ (e.g., AA �.�.��). Among the grammarians mentioned by Pān.ini there is another one to whom tradition ascribes a padapāt.ha, namely Gārgya. In the Sāmaveda, which occupies itself with the text of selected passages from the Rgveda on the level of syllables and larger metrical units, the position of the padapāt.ha is˚rather curious and seems to be, in the words of Burnell (����, xiv), “an accretion to the Sāmaveda (for the technical base of the gānas is the Saṁhitāpāt.ha).” Certain imperfections in this padapāt.ha which Burnell wants to understand as signs of its lateness, rather suggest that it belongs to an earlier date (than, for instance, Pān.ini), even if it was probably in some respects updated after Pān.ini.14 12 An example of the first meaning is found in RV �.���.��cd yád vā jágad jágaty ā ´hitaṁ padám “. . . and that the Jagat-line is based on the Jagat ˚ (viz., Jagat-hymn or chant).” In �� the poet continues his reflections but uses compounded -pad-instead of the word padá: gāyatrén.a práti mimı̄te arkám arkén.a sā ´ma traís..tubhena vākám / vākéna vākáṁ dvipádā cátus.padā-aks.áren.a mimate saptá vā ´n.ı̄h. // “According to the Gāyatrı̄ (-line) one makes the song of praise (arká); according to the song of praise a chant (sā ´ma), according to the Tris.t.ubh (-line) the recitation. According to the two- and four-lined recitation (one makes again a larger) recitation; according to the SYLLABLE they make the seven VOICES.” Cf. on these two verses Houben ����. In RV �.��.�� the compounded -pad- probably means another “foothold” in the recitation, the ˚ next smaller metrical unit after the line: the syllable (cf. Thompson ����, �-�). The wish to interpret padam in the series rcaṁ vārdharcaṁ vā pādaṁ vā padaṁ vā varn.aṁ veti (Kaus.ı̄taki- or Śāṅkhāyana-brāhman.a, KB ˚ ��.�) as ‘word’ (Renou ����, ��� [���]; Deshpande ����: iii) is understandable, but it is here more likely a ‘syllable’. On the passion for syllable counting in early Vedic literature (Brāhman.as): Jamison ����. 13 Deshpande remarks (����, ii footnote �): “Professor Cardona has drawn my attention to the difference between Geldner and Renou on this question. While Geldner thinks of ‘word’ as a metaphorical meaning of pada, Renou is more assertive in claiming that the term refers to ‘word’ in several RV passages.” Actually, the exact opposite is the case: Geldner (ad RV �.��.�) is ready to accept ‘word’ as a full-fledged meaning of Vedic padá and translates it occasionally ˚ as such, whereas Renou accepts ‘word’ for padá at the most as a metaphorical meaning. Renou ����, ���-��� [���-���] “Il est vrai que Geldner, dans son souci de moderniser le Veda, a soutenu que l’acception était déjà établie dans le RV I �� �, à la faveur d’un double sens (cf. la note ad loc. . . . ). C’est fort peu probable: il n’y a pas lieu de dissocier ces quelques phrases des nombreux passages où le terme désigne en contexte plus ou moins mystique le « séjour » de la divinité.” 14 Benfey ����, LVIIff.: the Sāmaveda padapāt.ha treats iva as a seperate word whereas both the Śākalya padapāt.ha of the Rgveda and the Pān.inian tradition (Vt. � on AA �.�.�) treat it as part of a compound. The˚Sāmaveda padapāt.ha splits words as compounds which Śākalya’s padapāt.ha presents as indivisible units, for instance puttra > put-tra; mitra > mi-tra and sūrya > su-urya/su-ūrya. Renou ����, ���, “Le padap. le plus primitif, donc sans doute le plus ancien, est celui du SV.” According to the Rgveda-prātiśākhya (RPr ed. Müller p. XIV) Gārgya wants words that end in a consonant to˚have the non-aspirated ˚ voiced phoneme of its class (e.g., vāg), but another grammarian, Śākat.āyana the non-aspirated non-voiced one (e.g., vāk ). The position elsewhere (Nirukta �.�) ascribed to Gārgya, that the prefixes have a meaning 169 Jan E.M. Houben �.�. The earliest Vedic text that shows a clear awareness of the word as linguistic unit, curiously still without reserving a distinctive term for it, is the Aitareya-āran.yaka (AiĀ), which explores in its third chapter the joining and disjoining of words. Śākalya is mentioned here among various teachers. The Aitareya-āran.yaka is at the same time the text which shows for the first time an interest in the padapāt.ha. The third chapter, also known as Saṁhitopanis.ad (with saṁhitā in the sense of Pān.ini’s AA �.�.��� equivalent to sandhi ), investigates, among other things, the implications of the coalescence of words (more precisely, of the final phoneme of the first word and the first phoneme of the second word) both from a cosmic (adhidaivam) and from a personal (adhyātmam) perspective. Its first lines are as follows: athātah. saṁhitopanis.at / prthivı̄ pūrvarūpaṁ dyaur uttararūpaṁ vāyuh. saṁ- ˚ṁhitetyasya māksavyo vedayāṁ cakre / sa hāvi- hiteti mān.d.ukeya ākāśah. sa . parihrto mene na me ’sya putren.a samagād iti samāne vai tat parihrto mena ˚ ˚ ity āgastyah. samānaṁ hy etad bhavati vāyuś cākāśaś ca / (AiĀ �.�.�) The passage may be translated as follows (following Keith except for an adaptation in the first sentence; my explanations added in square brackets): Next comes the Upanis.ad of the Saṁhitā [euphonic coalescence of words, sandhi ]. The former half [last phoneme of the first word] is the earth, the latter half [first phoneme of the second word] the heaven, their union [saṁhitā, i.e. sandhi ] the air, says Mān.d.ukeya. The union is ether, so proclaimed Māks.avya. ‘For it is not considered independent, and so I do not agree with his (Mān.d.ūka’s) son’, he said. ‘They are alike and it is considered independent’ said Āgastya: for the air and the ether are both alike. (Keith ����, ���) In this brief passage, which is characteristic for much that follows in this chapter, three things are remarkable. First of all, there is considerable disagreement between authorities. In a text dealing with the same subject and that is probably somewhat later, the Śı̄ks.ā-vallı̄ of the Taittirı̄ya-upanis.ad, the teaching is much more streamlined: the one who studies this Śı̄ks.ā-vallı̄ is not confronted with all sorts of disagreements but with a homogeneous, even if hardly less obscure, doctrine. Secondly, the domain of the coalescence and analysis of words inspires creative thinking. Thirdly, younger representatives of the tradition are confidently giving their own deviating view and defend it with a (sort of) argument. From this and other passages in the third or Saṁhitā-upanis.ad chapter of the Aitareya-āran.yaka we can only conclude that the analyzing and joining of words that link the word-by-word and the continuous recitation (padapāt.ha and saṁhitāpāt.ha), and which apply alternately in the intermediate kramapāt.ha, was at that time an exciting, entirely new subject. In the Aitareya-āran.yaka the saṁhitā-, pada- and krama-pāt.ha are referred to, respectively, as nirbhuja ‘twined’15 , pratrn.n.a ‘torn up’, and ubhayamantaren.a ‘in between both’, names that went out of use but ˚ were still remembered in the Rg- ˚ veda-prātiśākhya (RPrāti ed. Müller p. III, verse �.�-�). The Aitareya-āran.yaka clearly ˚ of their own (they are not just “illuminators” of a meaning already expressed), matches the practice in the SV padapāt.ha to present a prefix such as sam in samudram as an indepedent form (Scharfe ����, ��). 15 Madhav Deshpande ���� discusses these terms in his Introduction. For nirbhuja he suggests (p. V) ‘partless, uncut’, a meaning that is contextually justified but which does no justice to the evidently underlying root bhuj ‘bend’. 170  From Fuzzy-Edged “Family-Veda” isolates the linguistic unit ‘word’ with iti or sometimes iti vyāhrtih.,16 that is, with the ˚ general term for ‘utterance’. At AiĀ �.�.� pada may appear to mean ‘word’ although the meaning ‘line, quarter of a verse’ can here easily be maintained17 so that it has the same meaning as elsewhere in the AiĀ (�.�.�: nyūnāks.are prathame pade viharati, Keith: “He extends the first two verses [JH: lines of a verse] by a syllable”). Between the Rgvedic padá which refers at the most to a metrical unit and the grammarians’ pada as ˚ linguistic unit there is an undeniable gap which has attracted the attention of several philologists who have, however, not been able to convincingly bridge it. The semantic and conceptual gap remains: it points to a period of creative thinking such as reflected in the Aitareya-āran.yaka. The Aitareya-āran.yaka also happens to be the earliest Vedic text containing a passage that possibly refers to writing in the terms nollikhya, nāvalikhya. The two expressions refer to conditions which should not apply to a student and to activities which a student should not do if he is to receive instruction in the Mahānāmnı̄-verses. The limitations do not concern the beginner but the student who has already finished a large part of his study and who has probably already returned home and taken up new activities there before coming back for his concluding, advanced instructions (Falk ����). The expressions nollikhya, nāvalikhya come at the end of an enumeration of limitations which starts with precluding the instruction to those who have not (formally) become a student and who have not yet studied for a minimum period of a full year. Harry Falk (����) studied the terms and concluded (p. ��) that they have no link with writing, as is the interpretation of Sāyan.a, but that they rather refer, respectively, to “scratching the earth (in connection with going to toilet)” and the “shaving of the hair of the head” (“das Aufritzen des Erdbodens” and “das Rasieren des Haupthaars”). For these interpretations of the two terms, Falk brought together a number of places where they indeed have these meanings. A serious drawback of this interpretation, however, is that it renders the enumeration entirely haphazard.18 As Falk himself pointed out, there is also a text – although a post-Vedic one of uncertain age – where the two terms do have meanings in connection with writing: the Arthaśāstra. If the Aitareya-āran.yaka, as most of the Vedic texts, originated in the west and if it can be attributed to the �th century BCE, its authors must have been geographically near, even if they remained culturally distant, to areas where writing, and even (near-) alphabetic writing, had at that time been introduced and was practiced quite intensively, at least in some circles: the easternmost provinces of the Persian empire, including Gandhāra where the two cultures overlapped geographically. It can then not be excluded that the couple of terms, ullikhya and avalikhya, refers to the writing and erasing of writing on a wooden writing board, which was a common form of writing represented also in early Indian and (a little earlier) in Persian art. 16 AiĀ �.�.� sarve vedāh. sarve ghos.ā ekaiva vyāhrtih. prān.a eva “all Vedas, all sounds, are one utterance (‘word’), namely prān.a.” ˚ 17 AiĀ �.�.�, around the middle of the paragraph: atithim iti padaṁ bhavati. Instead of Keith ����: “. . . because there is in it [JH: in the hymn] the word ‘guest’,” it is rather to be translated as: “this is a metrical line (pada) containing ‘guest’.” 18 The interdiction to visit a barber (na nāpitena kārayitvā) and to take a bath (na snātvā), where Falk’s proposed interdictions would perhaps have fitted in, are found several places earlier in the list, before “having put on a flower-garland” (na srajam apinahya) and “approaching a woman” (na striyam upagamya). 171 Jan E.M. Houben �.�. The process of creating a pada-text to accompany (and probably to “edit”) an existing saṁhitā-text was known to Pān.ini and Patañjali as an ongoing human activity in contrast with the creation of the saṁhitā-text itself which is ārs.a ‘derived from Vedic seers’. It has been argued that the development and employment of (near-) alphabetic writing (old-Persian cuneiform Aramaic) was one of the pillars of the succesful administration of the Persian empire. The new invention can hardly have passed entirely unnoticed to Brahmins such as those living in Gandhāra. An initial confrontation of Brahmins with the practice of writing (in the Persian empire, including Gandhāra) even if it was not accepted in the heart of their own circles would in any case be an excellent candidate for having been a major stimulant for the creative thinking attested in the Aitareya-Āran.yaka, which started the development of a sophisticated technique of text transmission that could stand on its own feet and that enabled these circles to stay far from active participation in writing for several centuries. It is to be noted that, as the Vedic padapāt.ha marks the division into words and analyzes the mutual phonetic influence of these words, it does in this exactly what is to be done if speech or a continuous text is to be written down in a script that marks word boundaries and gives phonetic details of these words in a (near-) alphabetic script. A possible scenario for the discovery of the usefulness of a pada-analysis and later of the development of a padapāt.ha in the transmission of Vedic texts would then be: (�) Brahmins are aware of difficulties in transmitting central Vedic texts from the older to the younger generation. (�) some of them notice the surprising efficiency and reliability of the transmission of messages between officials of the Persian Empire, e.g., from the center of the Persian empire to Gandhāra, and vice versa. This message transmission (from official � to official �) occurs through the following steps: (a) complete oral message (from official �) > (b) the message is divided into words (scribe of official �) > (c) message written down in words and details are noted on phoneme level in the (near-) alphabetic script (scribe of official �) > (d) reading of the single words (reader official �) > (e) reconstruction and pronunciation of the continuous message to official � by his reader. Though these steps seem trivial to us who are learning to read and write at a very young age, they were not so at the time when near-alphabetic scripts were newly invented. (�) In the Vedic and Brahmanic milieu which is aware of writing but avoids it, it is nevertheless noticed that the activity of dividing a continuous message into words and the reconstruction of a continuous statement on the basis of the separated words contributes to the reliability of transmission (in contradistinction to a messenger who learns the message to be transmitted by heart, or someone who merely transmits the gist of a message). (�) However, given the aversion to writing, representatives of the Vedic tradition skip step (c), the actual writing of words. 172  From Fuzzy-Edged “Family-Veda” (�) Step (a) > (b) is developed for the Vedic hymns: it corresponds to the work of a pada-kāra, the author of a padapāt.ha on the basis of a continuous version of the existing sacred hymns; for step (d) > (e) rules are developed which combine the individual words in the padapāt.ha: this step corresponds to the work of another type of auxiliary text of the Veda, the Prātiśākhya. The development and use of a textual fixation without (physical) writing in the Vedic tradition confirms the position of J. Derrida (����, ����) – the character of “fixity” attributed to “writing” in opposition to “orality” inheres in language even if it is not (yet) physically written – vis-à-vis the objections of J. Goody (����). On the other hand, the (near-) alphabetic writing of the Persian empire seems to have inspired or incited the development of an alternative, purely oral way of “writing” which to some extent matches Goody’s argument for the crucial importance of the alphabet, although not as squarely as Goody would have wished. As even the entry of Alexander the Great in western India was almost completely “forgotten” in Indian traditions, it is not surprising that no record is available of the confrontation of the Brahmins of this region with the Persian empire, its bureaucracy, and its scripts. The earliest experiments with the pada-analysis have left no historical and even no narrative trace. But once the usefulness of pada-analysis was noticed in unambiguously establishing a text which had evoked doubts regarding some of its words, it was developed into the technique of pada-plus-saṁhitā recitation which facilitated the reliable transmission of large texts. If this made the learning of one’s family’s Veda more efficient, it must also have facilitated the learning of an additional Veda beyond the one traditionally studied in one’s own family. Bright students could now go on to become dvivedin or even trivedin. It also favoured the development of linguistic disciplines, espe� cially grammar (vyākaran.a). The early development of the pada-plus-saṁhitā recitation can therefore be regarded as an important landmark, as a turning point in the history of Vedic schools. It must have fortified the transmission in those schools which developed very early, pre-Pān.inian padapāt.has (Gārgya for the SV, Śākalya for the RV), and have ˚ inspired other, post-Pān.inian authors to do the same for other schools (Śaunakı̄ya AV, Ātreya for the TS). Could the creation of a padapāt.ha for the AV-Ś have contributed to its success in subsequent centuries even if in Pān.ini’s and Patañjali’s time it was rather the PaiS, without padapāt.ha, that was more prominent? Another side of the coin of the increased capacity of a school that adopts the pada-plus-saṁhitā technique is that it can possess and transmit a more extensive auxiliary literature. �.�. Since the creation of the padapāt.ha for various textual traditions was considered an activity for which human authors were remembered, we have to assume there was a time, not very long before Gārgya and Śākalya who were mentioned by Pān.ini, that the padapāt.ha had been neither created nor developed. We also have to assume that the transmission was less rigidly fixed before the introduction of the pada-plus-saṁhitā than afterwards. This means that we cannot follow Staal here that the Veda was “codified” at “around ���� B.C.E.” and that it then “entered the period of oral transmission . . . which fixed it for more than ���� years” (Staal ����, ��). Before the introduction of the pada-plus-saṁhitā the transmission was less rigidly fixed but even then not entirely haphazard. A reliable transmission was needed for the ritual, which was the main purpose for the transmission of the texts. As argued elsewhere (Houben ����, ����), ritual itself was a medium which transferred different types of 173 Jan E.M. Houben messages in two broad categories, canonical and performative ones. The former category is here relevant and it would obviously profit most from a very exact and stable transmission of texts to be employed in the rituals. A fascination with self-referential ritual patterns (Houben ����, ����), matching with self-referential patterns in Vedic poetry (Gonda ����, Thompson ����), must have contributed to a further stabilization of the ritual. Vedic ritual defined the context and the purpose but not the exact technique of the transmission. Can anything be said about this technique in very early times except for the non-use of writing and of pada-plus-saṁhitā recitation? It so happens that in a famous passage in the Rgveda the learning of Vedic recitation is used in a detailed analogy with ˚ older ones, who engage in continuous croaking at the onset of the rainy frogs, younger and season: akhkhalı̄kŕtyā pitáraṁ ná putró anyó anyám úpa vádantam eti ˚ as a son (approaches) the father (who speaks, that is, recites), transforming (the text) in (badly pronounced) syllables, one (of the frogs) approaches the other (frog) who speaks (that is, croaks). (RV �.���.�cd) ˚ It is here irrelevant whether, finally, the frogs illustrate (satirically) the behaviour of the Brahmins or the Brahmins the behaviour of the frogs, and also whether the hymn as a whole is meant to be “humorous” or a “serious” rain charm (Gonda ���� defending the second of each pair of positions). In any case, Thieme’s proposal (����, ���) – that akhkhalı̄kŕtya (padapāt.ha; for saṁhitāpāt.ha akhkhalı̄kŕtyā) refers to a faulty pronunciation by the˚son- student of the word aks.ára ‘syllable’ –˚suits perfectly the two situations that are being compared: the frogs that make noise at the beginning of the rains and the Brahmins who begin to study (also at the beginning of the rainy season). By contrast, the interpretation of Sāyan.a – widely accepted before Thieme’s proposal – takes akhkhalı̄kŕtya as a derivation from a supposed onomatopoeia (śabdānukaran.a) of akhkhala (cp. Wackernagel ˚ ����, �, who gives akhkhalı̄kŕtyā ‘jauchzend’ as the first example of “ ‘Onomatopoetische’ ˚ Nachbildungen von Naturlauten”), which would hardly be appropriate with reference to the Brahmins. To Thieme’s proposal must be added that there is no reason not to take final a > ı̄ before kr as significant, although we have here the only example in the Rgveda of this formation, ˚ which Pānini termed cvı̄ and which becomes more common in . ˚ the Taittirı̄ya-saṁhitā and especially in classical Sanskrit. Instead of “indem er Silben bildet” (Thieme) we should therefore interpret “transforming into (badly pronounced) syllables.” The fact that (according to much later witnesses) ks.a rather becomes kkha (not khkha) in Prakrit does not contradict that akhkhala would represent an imperfect pronounciation of aks.ára by the student-son in the Vedic period. If the comparison in the frog-hymn was appropriate for its own period – and there is no reason to assume it was not – the study of the Veda would indeed have been quite informal from a technical point of view, in the absence of both writing and a developed technique of pada-plus-saṁhitā recitation. Moreover, the hymn identifies the relationship between the young, learning Brahmin and the old, experienced one, as the relation between father and son. Even if “father” and “son” may include the situation where they are only “teacher” and “student”, the model-relation was that between father and son, which suggests that the teaching was primarily a family affair. Indeed, in earlier periods – presumably before the period of the creation of padapāt.has – the transmission of texts was organized first of all in families and clans, as is clear from the division into family books (in the Rgveda) which precedes a later organization of Rgvedic material into an integrated collection ˚ of ten ˚ 174  From Fuzzy-Edged “Family-Veda” man.d.alas. Within a family or clan, a well-established collection of hymns was transmitted to the younger generation, but there was also scope for brilliant young priest-poets to bring forward new hymns created in accordance with traditional poetic technique and to be employed at the next great ritual in order to make it more effective. The balance between stock examples to be learned and the scope for newly composed inspired hymns may have varied over time, in the direction of an increased emphasis on the established examples. As long as there was room for new compositions, the Veda was still “fuzzy-edged”. Before the development of the pada-plus-saṁhitā technique of text transmission enabled a more efficient study of larger amounts of texts – which also finally blocked any further creative development – we may assume that relatively small collections of hymns having sometimes specialized ritual functions were canonized within each specific family or group and remained in their custody. To this ancient situation which must have preceded the development of the pada-plus-saṁhitā technique of transmission we still find several references. One of the earliest is found somewhere in the middle of the ninth book of the Rgveda, as a whole devoted to Soma Pavamāna, at the end of a subgroup of hymns which ˚are all in the Gāyatrı̄ metre. In the last two verses of hymn �� which is the last of this subgroup, we find in a sort of ancient phala-śruti a reference to the advantages expected to fall to the share of the one who recites these verses, ŕcas, devoted to Soma Pavamāna: ˚ pāvamān´ı̄r yó adhyéty ŕ.sibhir sáṁbhrtaṁ rásam / tásmai sárasvatı̄ duhe ˚ks.ı̄ráṁ sarpír ˚mádhūdakam // The one who studies (recites) the verses devoted to Soma Pavamāna, the nectar collected by the seers, for him Sarasvatı̄ milks (provides) milk, ghee, honey and water. The existence of this group of �� hymns as a separate subsection is also evident from the additional hymn which follows hymn �� in the recently published Āśvalāyana-saṁhitā of the Rgveda. After enumerations of all types of sins which are purified by the recitation ˚ verses to Soma Pavamāna, verse �� gives a precise reference to the extent of this of the group of verses: dáśottarān.y 19 ´ni .sát. / dáśottarān.y rcā´ṁ caitát pāvamān´ı̄h. śatā ˚ etáj júhvañ jápaṁś caivá ghoráṁ mrtyubhayáṁ jayet// ˚ These are six hundred and ten verses to Soma Pavamāna; offering and reciting this group of verses, he should overcome the terrible fear of death. The number of six hundred and ten matches exactly the number of verses of hymns �-��. Hymn �� apparently reflects a development internal to the Āśvalāyana school which greatly expands the function very briefly suggested for the same group of hymns at the end of hymn ��. The hymn suggests the employment of this group of hymns in a Purān.ic type of ritual of moral purification, largely foreign to the ritual that underlies the hymns of this collection which presuppose their employment in a Soma-sacrifice. 19 Since the editor, as he explains in his Preface, has himself attributed accents and composed a padapāt.ha for original hymns (those not in the Śākala-saṁhitā), we need not hestitate to replace his dáśóttarān.y (which does not match his own padapāt.ha: dáśa-uttarān.i) by the expected accentuation of a bahuvrı̄hi: . 175 Jan E.M. Houben Another glimpse of the pre-padapāt.ha conditions of the transmission of Vedic texts we get in Paippalāda-saṁhitā �.��, which encourages solidarity among Brahmins when one of them is oppressed, and which contains an enumeration of various groups of Brahmins.20 All these groups of Brahmins are identified either according to the name of a Vedic seer from whom they apparently claim descent, or according to the hymns their clan or family transmits. The hymn thus contains references to (PaiS �.��.�) descendants of Jamadagni, of Kuśika, of Atri, of Kaśyapa, of Bharadvāja, to the Gotamas and to the Vasis.t.has; to (�.��.�) the Agastis, the Kan.vas, the Kutsas, the Prasravan.as, the Virūpas, the Gargas, the Mudgalas, the Yaskas, the Śaunakas and the Saṁkrtis; finally to (�.��.�21 ) Śatarcins, the great Mādhyama seers, the offspring of Ks.udrasūktas, ˚ i.e. authors of ksudrasūktas or . short hymns. The names Śatarcin, Mādhyama, Ks.udrasūkta all appear in AiĀ II.�.�-�, a narrative section in praise of prān.a ‘breathing’. Here it is clear from the context and from Sāyan.a’s commentary that reference is made to groups of Brahmins associated with various groups of hymns in the Rgveda: the Śatarcins, responsible for the groups of hundred verses which make up RV book ˚ �, the Mādhyamas comprising various groups responsible for the ˚ “middle books,” RV �-�, the Ks.udrasūktas who alone or together with the Mahāsūktas are responsible for ˚ book ��. The Grhyasūtras know partly parallel but more elaborate enumerations, e.g. ŚāṅkhGrS �.��.� ˚ in connection with a prayer due to deceased teachers in a ritual to be performed ˚ facing south and with upavı̄ta on the right shoulder: śatarcinah., mādhyamāh. – grtsamadah., viśvāmitrah., jamadagnih., vāmadevah., atrih., bharadvājah., ˚ vasis..thah., pragāthāh., pāvamānāh. – ks.udrasūkta-mahāsūktāh.. The references in AiĀ, in Grhya-sūtras such as ŚāṅkhGrS and in PaiS �.�� all refer to series of Vedic authors and ˚ offspring, but only in PaiS their ˚ they refer to Brahmins who are invoked to come to the rescue of a Brahmin suffering some misdoing. The PaiS hymn therefore adresses not just the spirits of the deceased but contemporaneous descendants who derive their name from their association with specific collections of hymns, presumably because they know how to employ those hymns in rituals and transmit them to succeeding generations. This situation would apply in a fullfledged way before the pada-plus-saṁhitā got established (a movement in which the Paippalāda apparently never actively participated), when the transmission of the Vedic texts was still primarily a matter of families. By contrast, the padapāt.ha provides a solid basis for the development of schools (rather than family-clans) competing within the same domain of ritual functionality (e.g. the Yajurveda), which differ only marginally, which can provide a strong formation even to students from different lineages. Vedic texts have been transmitted with much care and precision to subsequent gen� erations in the context of ritual, also in the pre-padapāt.ha period, when, however, even the best efforts left some scope for minor variation and deviation. Unexpectedly, we can 20 For this hymn in �� stanzas (� partly and �� entirely in prose): ed. D. Bhattacharya ����, ���-���; ed., translation and analysis A. Lubotsky ����. 21 Since the norm in book eight is the hymn of eleven verses (ekādaśarcah. kān.d.ah.), Lubotsky proposes to consider � and � as later additions. However, irrespective of questions of later or earlier, it should rather be considered that � (quarters a and b in prose) and �� (entirely in prose) have only a ritual justification and do not belong to the ŕcah. or verses of the hymn. The ˚ place of � does not point to its being a later insertion (as Lubotsky proposes), but it may have been the starting point of an originally independent hymn with similar ritual employment. If � is a later extension or an alternative in the ritual to �, PaiS �.�� would basically consist of two originally independent compositions of, resp., � and � ŕcah.. ˚ 176  From Fuzzy-Edged “Family-Veda” dispose of a quite detailed documentation of the outcome of these efforts. In a response to G. Possehl (����, ���), according to whom the long period of oral transmission of the Rgveda would leave so much scope for substantial change, H. Scharfe justly pointed out ˚ (����, ��-�� note ���) that the very early stability of the text of the Rgveda is confirmed by the “quotations in the traditions of other Vedic schools” such as˚the Sāmaveda and the ancient Brāhman.as, “with minimal deviations – and most of these deviations can be shown to be secondary.” The network of quotations and minor variations that substantiate Scharfe’s remark is analyzed in two voluminous works by Maurice Bloomfield, the Vedic Concordance (����) and Vedic Variants (����-����). Details in accent and phonetic variation are so minute that the overarching, long-term stability of the main texts is all the stronger confirmed. The variations noted in these studies must derive from developments before and even during the use of a padapāt.ha for some of the texts. �.�. The preceding exploration has shown that different modes of transmission of the Vedic texts and their rituals have quite distinct characteristics, that they are related to each other in a mostly irreversible sequence, and that they have implications for the character of the text. Although much further detailed research needs to be done, “knowledge transmission” appears to be an important parameter to add to the grid. With regard to a given text it will then be important to determine to which mode of transmission it (basically) belongs. Was a text composed at a time that writing was already widely accepted on the Indian subcontinent also for the transmission of Vedic texts? Or at a time that writing was unknown or at least avoided? Did it have a padapāt.ha to support its transmission before writing, and if so, when was that padapāt.ha produced? The logic of this parameter can be summarized as follows. (a) The mode of knowledge transmission has an impact on the character of a text. (b) Texts display characteristics which on close (philological) inspection reveal their authors as either “script-o-phobic” or as script-dependent. The former composes and initially transmits a text orally, the latter writes down his composition which he intends from the beginning to be for a public that is familiar with writing and written texts. Much research has been done on texts in the western tradition, research on a similar line with regard to classical and ancient Indian texts has started (Scharfe ����, �ff and references). (c) In between these two extremes there are intermediate and transitional stages (not counting for the moment conscious attempts to imitate, archaize): an author, e.g. Plato, participates already in a world of writing but the characters in his literary works, e.g., Socrates, are entirely oral or scriptophobic. (d) In a Vedic milieu the technique of pada-plus-saṁhitā transmission was developed, which could achieve, in a purely oral way, what (near-alphabetic) scripts achieved among their neighbours in the Persian empire (from around the �th cent. BCE). (e) The mode of knowledge transmission (oral in ritual context; pada- plus-saṁhitā; manuscript writing) had an impact on the development of Vedic schools which is visible in the internal arrangement of their texts and in their systematic mutual relations (complementary and competitive with reference to the ritual). 177 Jan E.M. Houben Adding the parameter of “knowledge transmission” to our previous Diagram, results in Diagram II. Diagram II: The Large Outlines Are Clear + “Knowledge Transmission” (�) RV, around ���� BCE (estimate): Vedic literature and Vedic ritual are at first orga� ˚ nized according to expansive RV-families and clans: each parallel and competitive ˚ group has a similar array of textual materials; / Area: NW of Indian subcontinent: larger Punjab / Transmission: RV-dominated ritual, family-clans, “frog-style”- transmission (RV �.���). ˚ ˚ (�) Late+Post-RV, after ���� BCE (estimate): Development of Yajurveda and of Śrauta-ritual˚ in which three, later four Vedas have a role: functional diversification and re-organization: some families focus (continue their focus) on Rg-recitation, others (acc. to gotra usually not different from RV-families) specialize ˚in Yajurvedic ritual knowledge, in Sāma-chanting, etc.; ˚ / Area: Kuruks.etra and environment / Transmission: family-clans in ritual-functional diversification, “frog-style”- trans� mission. (�) Late+Post-YV, from ��� BCE (estimate): Formation and development of special� ized and, per Veda, mutually only marginally differring schools; / Area: first the north, next the whole subcontinent. / Transmission: development PADA from �th BCE.; family-clans become schools, more open for non-family Brahmin-students outside family- clan; ritual-functional diversification but also competition within the same ritual functionality, esp. Ya� jurveda; PADA (stronger) schools and non-PADA (weaker) schools. �. The Large Outlines Plus Knowledge Transmission Plus Ritual �.�. To these large outlines based on parameters A, B and C, we now add a new parameter (D), that of “Ritual.” Vedic ritual, which, as we have seen, serves as context for the transmission of knowledge – first on its own and later together with other media or modes of knowledge transmission – has other aspects which make it interact with the environment and economy. The texts are linked to rituals in several ways, through explicit or indirect references, so that texts and sections of texts can be linked with a ritual; next, it is to be determined how the ritual presupposed in the text is related to ecological and economic reality as (potentially) evident in archaeological and, more specifically, paleoecological research. As for paleoecology, relatively little certified research results are available for the north of the Indian subcontinent (covering the area mainly of current Afghanistan, Pakistan and the north of India). To be more concrete, a study similar to Trees and Woodlands of South India: archaeological perspectives (Asouti & Fuller ����) which, as its title indicates, deals mainly with South India, should have been available for the north of the Indian subcontinent. In order for such a study to be possible for the north of the Indian subcontinent, more detailed researches should have been available that 178  From Fuzzy-Edged “Family-Veda” focus on problems of pastoralism and agro-pastoralism (rather than on settled agriculture) and their ecological environment (e.g. woodlands) for the period and the area concerned (rather than on the period and area of the Indus civilization or on neolithic pastoralism). It is of course possible to make a rather long list of somehow relevant studies22 – apart from the indirectly relevant Asouti & Fuller ����: Allchin ����; Gadgil & Guha ����; Meadow & Patel ����23 ; Chattopadhyaya ����; Fuller ����, ����; Ratnagar ���� – but their implications for the problems at stake are often far from evident. For this reason, I will here merely summarize the main points of my previous discussion (Houben ����), and add a summary of the logic underlying this parameter. �.�. The Vedic texts presuppose rituals, sometimes very specifically, almost like a liturgy which follows the ritual or part of the ritual step by step, for instance the Soma-hymns of the �th book of the Rgveda (Oberlies ����), the Rgvedic Āprı̄-sūktas (Bosch ����), ˚ and the Pravargya (Houben ����); often in a more˚generalized way, when the hymn consists of a kind of prologue to the real sacrificial ritual and serves to invite the divinity to be present (Geldner ����, ���; Renou ����, �); and sometimes the presupposed ritual context is indeterminate (Gonda ����). Also between classical (esp. Śrauta) Vedic ritual (as described in the somewhat later Śrauta-sūtras) and the Rgveda there is a very strong connection, even if there have been some transformations. In˚the case of the various Vedic schools of the Yajurveda and the Sāmaveda, the connection between the ritual and the texts is very direct (Gonda speaks of “liturgical Saṁhitās”: ����, ���-���). In the case of the Atharvaveda there is a connection with classical Śrauta, but more with Grhya and what can be called Atharvanic ritual. The main point is that the Vedic core texts,˚esp. the Saṁhitās, are strongly related to rituals, either partly known ones because of our lack of information (RV, AV), or quite well known (Yajurvedic and Sāmavedic texts). This is well-known and˚widely accepted. It is or should be equally well-known that the rituals of the various Vedic texts, starting with the oldest collection, the Rgveda, presuppose a certain mode of life, which implies a certain way of interacting with ˚ the natural but also the economic environment. It is since the beginning of modern Veda research well-known that the Rgveda and other Vedic texts, esp. the Yajurvedic Saṁhitās, contain references both to ˚ pastoralism and to agriculture. At present, referring to very extensive research of current and historical forms of pastoralism and agriculture this loose reference to modes of existence is not sufficient. We can and should define the mode of existence implied in the Rgveda and the rituals it implies with more precision as “agro-pastoralism”. The importance ˚ of the specific mode of existence of non-settled pastoralism has been highlighted by scholars such as J.C. Heesterman ���� and Bruce Lincoln ����. I follow here a discussion and summary of current insights by the archeologist Shereen Ratnagar in her article “Agro-pastoralism and the Migrations of the Indo-Iranians” that appeared in ����: Characteristics of agro-pastoralism (following Ratnagar ����): 22 Some relevant studies and discussions before June ���� can be found through Misra & Kanungo ����. 23 What the authors observe on the late Harappan Period is relevant for the beginning of the period that has our interest here. Their remark on p. ��� applies as well to later centuries in the same area: “What we have currently for South Asia in general and for the north-western portion in particular is a patchy archaeological framework with little depth of understanding for any single region, let alone for the entire area.” 179 Jan E.M. Houben (�) those who practice agro-pastoralism appreciate domestic animals much more than agricultural products, even if they may consume more grain products than livestock products (milk, meat etc.). (�) they invest the positive results of their crops in the increase of their herds. These two traits together are at the basis of a related phenomenon: (�) the tendency of agro-pastoralists to geographic expansion. On the prosaic topic of (�) the ancient Vedic texts do not provide positive information. Perhaps its validity can be inferred from the absence of references to storehouses for grain, etc.; indications for (�) and (�) are abundantly available. In a time when no sophisticated techniques were available to transmit a ritual, we can expect it to prosper and expand only in a supportive environment, when there is a positive match between ritual and environment. In the Rgvedic and early post-Rgvedic period, when use was made only of an akhkhalı̄kŕtyā-type ˚ of transmission and˚a transmission ˚ determined by ritual needs, such a positive match must have been there. The Rgveda and next the Yajurvedic texts and the ritual system they presuppose turn out to be˚ thoroughly agro-pastoral in character. Cattle occupy a most honorable position, offerings of freshly milked milk are recurrent, the ritual is basically mobile. Referring to the work of J.C. Heesterman ����, Bruce Lincoln ���� and Shereen Ratnagar I will take this here as a starting point. Now, agro-pastoralism is basically expansive in character and requires an ecological environment which it tends to transform in the course of time. These two data provide us with chronological and geographical parameters that can be matched not only with dispersed textual testimonies but also with paleoecological findings on the historical and pre-historical presence of woodlands and cultivated areas and on population ratios in the entire northern part of the Indian subcontinent, from Afghanistan via Kuruks.etra to the Gangetic plains and Magadha. As explained, more archaeological and paleoecological researches would be needed. Feasible investigations can be defined that can be expected to provide further crucial (paleoecological) evidence. The next major stage of ecological development observed in India is the transition to an agricultural and largely urbanized environment in (Greater) Magadha (cf. Bronkhorst ����), which coincides with the beginnings of Jainism and Buddhism in the sixth – fifth century BCE. The Vedic semi-moblile agro-pastoral grhapati has become the peasant, the settled, agricultural gahapati, omnipresent in Buddhist˚ texts (e.g., Chakravarti ����). The early development of the last of the major Vedic divisions to become independent, the Atharvavedic branches – which, unlike their Yajurvedic and Sāmavedic predecessors, do not get a chance to develop completely before they are almost entirely swept away by new ecological and religious developments – lays in between these two, and shows not only evidence of a largely agricultural environment, but also of a general resource crunch: Brahmins who have given up an earlier semi-nomadic life-style in favour of a more settled one are experiencing increased difficulties to find stable niches for survival. If early emerging AV-schools, including Paippalāda (mentioned by Patañjali in his Mahābhās.ya together with other well-known schools of his time) and an in earlier times little attested Atharvanic Śaunaka, went in the direction of settled agriculture first,24 24 The status attributed to the rice dish, brahmaudana in AV hymns and rituals (Gonda ����), is diametrically opposed to the agro-pastoralist tendency to value cattle more than agricultural 180  From Fuzzy-Edged “Family-Veda” away from the agro-pastoral main-stream canonizing schools of RV, YV and SV, it was Śaunaka’s tradition which curved back most energetically to an ˚ already canonized Śrauta-ritual system in which Rgvedins, Yajurvedins and Sāmavedins were participating and finding their employment. ˚ Perhaps it was only at this relatively late moment that the Śaunakas created and accepted a padapāt.ha for their tradition, unlike the Paippalādins who remained aloof and without padapāt.ha and whose text appears to have deteriorated in subsequent centuries. Although Śaunaka turned towards a canonized Śrauta-system, it was there not very warmly accepted (Gonda ����, ��� note ��). As is well-known, ritual handbooks of the Atharvaveda reserve the function of a Brahman to the Atharvavedin, but schools such as the Āpastambins do not support the claim and attribute this function to someone of the Rgveda, Yajurveda or Sāmaveda. Let us go back ˚ to the agro-pastoralism of the Rg- and Yajurveda which was basically ˚ expansive in character and required an ecological environment which it tends to transform in the course of time. We can now ask the question: when and where did they find a suitable ecological environment, and when did they transform that environment in the course of time? Going back in time from the time of the Buddha when Vedic agro-pastoralism was already outdated, we find a suitable match – in need of further research – somewhere in the centuries preceding the Buddha, so preceding the �th century BCE., when the northern half of the Indian subcontinent from west to east was presumably still rich in uncultivated areas, and when Vedic clans or tribes adopting “[a]n ethic of exhaustive resource use, with the [Yajña] as its cornerstone” (Gadgil & Guha ����, ��) found there extensive exploitable niches. It is tempting to formulate the thesis that areas in the north of the Indian subcontinent which the Vedic people of the late Bronze Age (RV) and other late neolithic populations (those responsible for the Deccan Ashmounds)˚had not been able to cultivate, could now be cultivated by mobile agro-pastoralist Vedic people of the early Iron Age (YV), and could subsequently be settled by more agriculturally oriented Vedic people of the early Iron Age (AV). Some textual evidence is provided, inter alia, by a passage of the Śatapatha-brāhman.a, �.�.�.��-�� (discussed in Houben ����) according to which a certain “Māthava, the Videgha,” having started from the Sarasvatı̄ river in the west, had come to the east, carrying Agni, the fire god, and crossed the Sadānı̄rā (current Gandak river) in order to settle, for the first time as Brahmin, east of it. The passage is often cited in order to point out an eastward expansion of Brahmanism. However, it has been overlooked that the passage also very clearly points to a transformation of the land from marshy and uncultivated to cultivated: tád dhā ´ks.etrataram ivāsa srāvítaram ivā ´svaditam agnínā vaiśvānarén.éti // tád u haitárhi / ks.étrataram iva brāhman.ā ´ u hí nūnám enad yajñáir ásis.vadan[t]. At that time it (the land east of the Sadānı̄ra) was very uncultivated, very marshy, because it had not been tasted by Agni Vaiśvānara. Nowadays, however, it is very cultivated, for the Brāhmans have caused (Agni) to taste it through sacrifices. In this ecological transformation from “very uncultivated, very marshy” land to “very cultivated” the ritually mastered Agni ‘Fire’ played a crucial role. The god of Fire also products, see above point (�). On the Paippalādins’ “closeness to the contemporary farmers” see Bhattacharya ����, liii. 181 Jan E.M. Houben played an important role in a curious episode of the Mahābhārata, the burning of the Khān.d.ava forest by Arjuna and Krs.n.a in which large numbers of forest-inhabiting animals and, according to one sentence in ˚ the critical edition, also humans, were said to have been killed. The event must have contributed to the cultivation of the area where the Pān.d.avas wanted to settle. It was interpreted in terms of a conflict between populations with different modes of resource use by Karve (����) and Gadgil & Guha (����). Still another transformation seems to have taken place at the confluence of the Gaṅgā and the Yamunā: this landmark was in one text characterized by the presence of a forest, the Kālaka forest, and in a later text by the presence of a city or settlement, Prayāga (see passages discussed in Houben ����). The Vedic ritual system must have been in an early stage of creative development and it catalyzed the transformation of uncultivated woodland into cultivated land suitable for pastoralism and agriculture. The Vedic people’s ritual and religion system had a progress-oriented character, to which Gadgil & Guha refer with a term from population ecology: r-strategist character. This refers specifically to the equation of population dynamics proposed by Pierre François Verhulst in ����. In this equation, r = growth rate and K = capacity of the environment to support a population. In the light of this equation, three types of biological organisms (and, by extension, three types of sociological “organisms”) can be distinguished: (a) r -strategist: strives for fast and massive reproduction (the limit is defined by r ); (b) K -strategist: niche-exploitation (limit defined by K ); (c) those having a continuous spectrum of r -traits and K -traits. Precisely thanks to the broad overall success of the semi-nomadic agro- pastoralism of Vedic people in the areas that are in transformation, the population grows rapidly and the earlier favourable land-to-man and livestock-to-man ratios decrease significantly. The society and its environmental context leave little scope for expansionist r -strategists. Instead, society enters a phase suitable for niche-exploiting K -strategists. The numerically strong components of the population engaged in by now settled agricultural pastoralism are in need of a new belief system that would stress careful and sustainable patterns of resource use. Buddhism and Jainism proved to be able to cater to this broadly-felt need. Brahmins as inheritors of the old r -strategist belief system opposed the new religions Buddhism and Jainism. They became marginalized, but thanks to their exceptional tenacity and efficient modes of textual transmission they were able to maintain their system against all odds for several centuries, until the general Pus.yamitra brought about a revival, a sort of Ur-revival, of Brahmanism. The favorable and creative “r -strategist” period of Brahmanism was over for good, and had given way to a time of “K-strategist” niche-exploiting. The mnemotechnic and intellectual skills developed by Brahmins in their young years proved useful in numerous other domains beyond the ritual. However, as far as the Vedic texts and rituals were concerned, the sophisticated techniques of knowledge transmission were allowing the memorization of impressive quantities of text, but blocked creative additions to the now completely closed canons. 182  From Fuzzy-Edged “Family-Veda” Map25 with approximate indication of the areas of Bactria, Arachosia and Gandhāra (provinces in the Persian empire in the �th -�th centuries BCE), the area of Kuruks.etra, the city Prayāga (Allahabad), and the river Sadānı̄rā (Gandhak), flowing from the Himālaya into the Ganges. Based on: Plate � “Orographical features”, Imperial Gazetteer, Atlas of India, ����. �.�. The logic of this parameter can be summarized as follows. (a) Ancient Vedic texts presuppose a ritual, either one about which details are known from other (later) sources, or one for which further details are missing. While numerous details may remain open for discussion, the large outlines of the rituals presupposed in the texts can be established. 25 I thank E. Giraudet for the production of the map. 183 Jan E.M. Houben (b) The ritual presupposed in a text implies an interaction with the environment and economy, in the case of ancient Vedic ritual, especially Rgvedic and Yajurvedic ritual: agro-pastoralism. (c) The ritual may match the actually available environment: this leads to the expansion of the ritual and of those engaged in the ritual. The ritual is “fuzzy-edged”, still open to creative expansion, (d) The ritual may mismatch the actually available environment: ritual canons of the four Vedas, the Catur-Veda, are closed, the search for exploitable niches has become necessary. (e) The philological characteristics of a text and its “r -strategist” or “K-strategist” mode can be determined and matched with archaeological and paleoecological research (desideratum!), in order to contribute to a further localization in time and space of early Vedic people and their texts in order to shed light on the two most obscure millennia in the (proto-) history of the Indian subcontinent: the first and second millennium BCE. Applying the new variable “Ritual” to the previous Diagram leads to a new Diagram III: Diagram III: The Large Outlines Are Clear + “Knowledge Transmission” + “Ritual” (�) RV, around ���� BCE (estimate): Vedic literature and Vedic ritual are at first orga� ˚ nized according to expansive RV-families and clans: each parallel and competitive ˚ group has a similar array of textual materials; / Area: NW of Indian subcontinent: larger Punjab. / Transmission: RV-dominated ritual, family-clans, “frog-style”- transm. (RV �.���). ˚ ˚ / Ritual: Text and ritual are in an expansive phase, “fuzzy-edged” Veda; the Vedic people are in an r -strategist phase, agro-pastoralist expansion. (�) Late+Post-RV, after ���� BCE (estimate): Development of Yajurveda and of Śrauta-ritual˚ in which three, later four Vedas have a role: functional diversification and re-organization: some families focus (continue their focus) on Rg-recitation, others (acc. to gotra usually not different from RV-families) specialize ˚in Yajurvedic ritual knowledge, in Sāma-chanting, etc.; ˚ / Area: Kuruks.etra and environment / Transmission: family-clans in ritual-functional diversification, “frog-style”- transm. / Ritual: Texts (esp. RV) more and more canonized, ritual (YV) expansive; Vedic ˚ people in strong agro-pastoralist expansion, or tend to agricultural settlement (AV). (�) Late+Post-YV, from ��� BCE (estimate): Formation and development of specialized and, per Veda, mutually only marginally differring schools; / Area: first the north, next the whole subcontinent. 184  From Fuzzy-Edged “Family-Veda” / Transmission: development PADA from � th BCE.; family-clans become schools, more open for non-family Brahmin-students outside family-clan; ritual-functional diversification but also competition within the same ritual functionality, esp. Ya� jurveda; PADA (stronger) schools and non-PADA (weaker) schools. / Ritual: Vedic texts and rituals are canonized (Catur-Veda), Vedic people enter K -strategist phase, Brahmins need land rather than cows. �. Conclusion The formation of the Veda and the development of Vedic schools in ancient India constitute a complex phenomenon that is, from a global perspective, entirely unique in character and extent, even if components of this complex phenomenon can play important roles in the scientific study of what has recently been termed “natural experiments of history”. Quod Erat Demonstrandum. 185  References AA = Pān.ini’s As.t.ādhyāyı̄ (as..ta+adhyāya+Ṅı̄P acc. to AA �.�.��). AiĀ = Aitareya-Āran.yaka; ed. and tr. by A.B. Keith, Oxford ����. Allchin, Raymond F. ����. Neolithic Cattle Keepers of South India: A study of the Deccan Ashmounds. Cambridge: Cambridge University Press. Asouti, Eleni and Dorian Q. Fuller. ����. Trees and Woodlands of South India: Archaeological Perspectives. Walnut Creek, CA: Left Coast Press. AV = Atharva-Veda (Śaunakı̄ya). BaudhDS = Baudhāyana-Dharma-Sūtra. Benfey, Theodor. ����. Die Hymnen des Sâma-Veda. Leipzig: F.A. Brockhaus. Bhattacharya, Dipak. ����. The Paippalāda-Saṁhitā of the Atharvaveda. Vol. One. Consisting of the first fifteen Kān.d.as. Calcutta: The Asiatic Society. ——— . ����. The Paippalāda-Saṁhitā of the Atharvaveda. Vol.Two, Consisting of the sixteenth Kān.d.a. Calcutta: The Asiatic Society. Bloomfield, Maurice. ����. The Atharvaveda. Strassburg: Karl J. Trübner. ——— . ����. Vedic Concordance. Being an alphabetic index to every line of every stanza of the published Vedic literature and to the liturgical formulas thereof, that is an index to the Vedic mantras, together with an account of their variations in the different Vedic books. Cambridge, Mass.: Harvard University. [See now also: Franceschini ����.] ———, and Franklin Edgerton (vols. �-�); vol � in addition: M.B. Emeneau. ����-����. Vedic variants: a study of the variant readings in the repeated mantras of the Veda. Philadelphia: Linguistic Society of America. Bosch, Lourens van den. ����. “The Āprı̄ hymns of the Rgveda and their interpretation.” Indo-Iranian Journal ��: ��-���, ���-���. ˚ Bronkhorst, Johannes. ����. “Some observations on the Padapāt.ha of the Rgveda.” Indo-Iranian Journal ��: ���-���. ˚ ——— . ����. “L’indianisme et les préjugés occidentaux.” Etudes de Lettres, avril-juin ����: ���-���. ——— . ����. “Literacy and Rationality.” Asiatische Studien / Études Asiatiques ��.�: ���-���. ——— . ����. Greater Māgadha. Leiden: E.J. Brill. Bryant, Edwin F. and Laurie L. Patton. ����. The Indo-Aryan Controversy: Evi� dence and Inference in Indian History. London: Routledge. ���  From Fuzzy-Edged “Family-Veda” Burnell, A.C. ����. The Sâmavidhânabrâhman.a (being the third Brâhman.a) of the Sâma Veda. Text and commentary, with Introduction. London: Trübner & Co. Caland, Willem. ����. “Ueber das Vaitānasūtra und die Stellung des Brahman im Vedischen Opfer.” Wiener Zeitschrift zur Kunde des Morgenlandes, ��: ���-���. [Caland ����: ���-���.] ——— . ����. “Zur Atharvavedalitteratur.” Wiener Zeitschrift zur Kunde des Morgen� landes, ��: ���-���. [Caland ����: ���-���.] ——— . ����. Willem Caland: Kleine Schriften. Ed. by M. Witzel. Stuttgart: Franz Steiner. Chakravarty, Uma. ����. Every Day Lives, Every Day Histories: Beyond the Kings and Brahmanas of ‘Ancient’ India. New Delhi: Tulika Books. Chattopadhyaya, U.C. ����. “Researches in Archaeozoology of the Holocene Period (including the Harappan Tradition in India and Pakistan).” Indian Archaeology in Retrospect, Volume III: Archaeology and Interactive Disciplines (ed. S. Settar and R. Korisettar): ���-���. Delhi: Manohar / Indian Council of Historical Research. Collins, Randall. ����. Sociology of Philosophies: A Global Theory of Intellectual Change. Cambridge, Mass.: Belknap Press of Harvard University Press. Derrida, Jacques. ����. L’Ecriture et la Différence. Paris: Editions Seuil. ——— . ����. Writing and Difference (English transl. by Alan Bass). London: Routledge & Kegan Paul. Deshpande, Madhav M. ����. Recitational Permutations of the Śaunakı̄ya Athar� vaveda: Critically Edited with an Introduction. Cambridge, Mass.: Harvard University, Dept. of Sanskrit and Indian Studies. Diamond, Jared, and James A. Robinson. ����. Natural Experiments of History. Cambridge, Mass.: Belknap Press of Harvard University Press. Eisenstein, Elizabeth L. ����. The Printing Press as an Agent of Change: communi� cations and cultural transformations in early modern Europe. Cambridge: Cambridge University Press. Falk, Harry. ����. “Goodies for India – Literacy, Orality, and Vedic Culture.” In Erscheinungsformen kultureller Prozesse (ed. by W. Raible): ���-���. Tübingen: Gunter Narr Verlag. ——— . ����. “AĀ �.�.�.: nollikhya nāvalikhya.” Indo-Iranian Journal ��: �-��. ——— . ����. Schrift im alten Indien. Ein Forschungsbericht mit Anmerkungen. Tübingen: Gunter Narr Verlag. Franceschini, Marco. ����. An updated Vedic concordance: Maurice Bloomfield’s A Vedic Concordance enhanced with new material taken from Vedic texts. Cambridge, Mass.: Harvard University. 187 Jan E.M. Houben Fuller, Dorian Q. ����. “Fifty Years of Archaeobotanical Studies in India: Laying a Solid Foundation.” Indian Archaeology in Retrospect, Volume III: Archaeology and Interactive Disciplines (ed. S. Settar and R. Korisettar): ���-���. Delhi: Manohar / Indian Council of Historical Research. ——— . ����. “Non-human genetics, agricultural origins and historical linguistics in South Asia.” The Evolution and History of Human Populations in South Asia (ed. by M.D. Petraglia and Bridget Allchin): ���-���. Dordrecht: Springer. Gadgil, Madhav, and Ramachandra Guha. ����. This Fissured Land: An Ecological History of India. Delhi: Oxford University Press. Geldner, Karl Friedrich. ����. “Rgveda �.��.” Vedische Studien II: ���-���. ˚ ——— . ����. Der Rig-Veda aus dem Sanskrit im deutsche übersetzt und mit einem laufendem Kommentar versehen. Vols. �-�. Cambridge, Mass.: Harvard University Press. Gonda, Jan. ����. “The so-called secular, humerous and satyrical hymns of the Rgveda.” Reprinted in Selected Studies vol. III: Sanskrit: Grammatical and Philolog� ˚ Studies: ���-���. Leiden: E.J. Brill, ����. ical ——— . ����. The Vision of the Vedic Poets. Leiden: E.J. Brill. ——— . ����. The Savayajñas (Kauśikasūtra ��-��) ; Translation, Introduction, Commentary. Amsterdam: North-Holland Publishing Company. ——— . ����. Vedic Literature. Wiesbaden: Harrassowitz. ——— . ����. Hymns of the Rgveda not employed in the Solemn Ritual. Amsterdam: ˚ North-Holland Publishing Company. Goody, Jack. ����. The Domestication of the Savage Mind. Cambridge: Cambridge University Press. ——— . ����. “Oral composition and oral transmission: the case of the Vedas.” In Oralità: Cultura, Letteratura, Discorso (ed. by B. Gentilli and G. Paioni): �-��. Roma: Edizioni dell’Ateneo. ——— . ����. The Interface Between the Written and the Oral. Cambridge: Cambridge University Press. The Interface Between the Written and the Oral. Cambridge: Cambridge University Press. ——— . ����. The Power of the Written Tradition. Washington and London: Smith� sonian Institution Press. Goody, Jack and Ian Watt. ����. “The Consequences of Literacy.” Comparative Studies in Society and History �.�: ���-���. Havelock, Eric A. ����. The Liberal Temper in Greek Politics. New Haven: Yale University. ——— . ����. Preface to Plato. Cambridge, Mass.: Harvard University. 188  From Fuzzy-Edged “Family-Veda” ——— . ����. The Muse Learns to Write: Reflections on Orality and Literacy from Antiquity to the Present. New Haven: Yale University Press. Heesterman, J.C. ����. The Broken World of Sacrifice: An Essay in Ancient Indian Ritual. Chicago: Chicago University Press. Hinüber, Oskar von. ����. Der Beginn der Schrift und frühe Schriftlichkeit in Indien. Stuttgart: Franz Steiner. Houben, J.E.M. ����. “The ritual pragmatics of a Vedic hymn: The ’riddle hymn’ (Rgveda �.���) and the Pravargya-ritual.” Journal of the American Oriental Society ˚ ���.�: ���-���. ——— . ����. “Rgveda �.���.��-�� and Bhartrhari’s Philosophy of Language.” In Expanding and˚Merging Horizons: Contributions ˚ to South Asian and Cross-Cultural Studies in Commemoration of Wilhelm Halbfass (ed. by K. Preisendanz): ���-���. Wien: Verlag der Österreichischen Akademie der Wissenschaften. ——— . ����. “Transmission Sans Ecriture dans l’Inde Ancienne: Enigme et Structure Rituelle.” In Ecrire et Transmettre en Inde Classique (ed. Colas, Gérard, and Gerdi Gerschheimer): ��-���. Paris: Ecole Française d’Extrême Orient. ——— . ����. “Formal Structure and Self-referential Loops in Vedic Ritual.” In Ritual Dynamcis and the Science of Ritual, Section I: Grammar and Morphology of Ritual (ed. by Axel Michaels and Anand Mishra): ��-��. Wiesbaden: Harrassowitz. ——— . ����. “Vedic ritual as medium in ancient and pre-colonial South Asia: its expansion and survival between orality and writing.” In Travaux de Symposium International: Le Livre. La Roumanie. L’Europe. Troisième édition – ��-�� Sepembre ����. Tome III: Section IIIA: Le Veda-Vedāṅga et l’Avesta entre oralité et écriture (ed. by J.E.M. Houben & J. Rotaru): ���-���. Bucarest: Bibliothèque de Bucarest. ——— . ����. “Les Perfectibles (sādhyá) entre circularité et causalité du rituel védique.” La Clairière: Etudes pour Charles Malamoud. Ed. by Silvia D’Intino and Catherina Guenzi: ��-��. Turnhout: Brepols. Houben, Jan E.M., and Saraju Rath. ����. “Manuscript Culture and its impact in ‘India’: Contours and Parameters.” In Saraju Rath (ed.), Aspects of Manuscript Culture in South India: �-��. Leiden: E.J. Brill. Hudson, N. ����. “Challenging Eisenstein: Recent Studies in Print Culture.” Eigteenth- Century Life, Volume ��, no. �: ��-��. Johns, Adrian. ����. The Nature of the Book: Print and Knowledge in the Making. Chicago: University of Chicago. Jamison, Stephanie. ����. “Brāhman.a syllable counting, Vedic tvác ‘skin’, and the Sanskrit expression for the canonical creature.” Indo-Iranian Journal ��: ���-���. Karve, Irawati. ����. Yuganta: the end of an epoch. �nd , rev. edition. New Delhi: Sangam Books. [�st ed. Poona: Deshmukh Prakashan, ����.] 189 Jan E.M. Houben KB = Kaus.ı̄taki- or Śāṅkhāyana -Brāhman.a of the Rgveda. ˚ Keith, Arthur Berriedale. ����. The Aitareya Āran.yaka: ed. from the Mss. in the India Office and the Library of the R. Asiatic Soc. with introduction, translation . . . and an appendix containing the portion hitherto unpublished of the Śāṅkhāyana Āran.yaka. Oxford: Clarendon Press. Lincoln, Bruce. ����. Priests, warriors, and cattle: a study in the ecology of religions. Berkeley: University of California Press. Lubotsky, Alexander. ����. “PS �.��. Offense against a Brahmin.” In The Atharvaveda and its Paippalādaśākhā: Historical and Philological Papers on a Vedic Tradition (ed. A. Griffiths and A. Schmiedchen): ��-��. Aachen: Shaker Verlag. Meadow, Richard H. & Ajita K. Patel. ����. “From Mehrgarh to Harappa and Dholavira: Prehistoric Pastoralism in North-Western South Asia through the Harap� pan Period.” Indian Archaeology in Retrospect, Volume II: Protohistory: Archaeology of the Harappan Civilization (ed. S. Settar and R. Korisettar): ���-���. Delhi: Manohar / Indian Council of Historical Research. Misra, V.N. and A.K. Kanungo. ����. A Bibliography of Indian Archaeology: Prehistory, Protohistoric and Early Historic Periods. [Index of publications up to June ����] New Delhi: Archaeological Survey of India. Oberlies, Thomas. Die Religion des Rgveda. Zweiter Teil – Kompositionsanalyse der Soma-Hymnen des Rgveda. Wien:˚Insitute of Indology, University of Vienna. ˚ Oldenberg, Hermann. ����. Metrische und textgeschichtliche Prolegomena zu einer kritischen Rigveda-Ausgabe. Berlin: Hertz. Olivelle, Patrick. ����. Dharmasūtras: The Law Codes of Āpastamba, Gautama, Baudhāyana and Vasis..tha. Annotated Text and Translation. Delhi: Motilal Banarsi� dass. ——— . ����. Manu’s Code of Law. A critical edition and translation of the Mānava-Dharmasūtra. New York: Oxford University Press. Ong, Walter J. ����.Orality and Literacy: The Technologizing of the Word. London: Methuen. PaiS = Paippalāda-Saṁhitā. Vol. �: Consisting of the first fifteen Kān.d.as. ed. D. Bhattacharya. Calcutta: The Asiatic Society, ����. Possehl, Gregory. ����. Review of The Indo-Aryans of Ancient South Asia (ed. Erdosy ����), Berlin: Walter de Gruyter. Journal of the American Oriental Society ���: ���-���. Rath, Saraju. ����. “Vedic education in early mediaeval India according to North Indian Charters.” In Travaux de Symposium International: Le Livre. La Roumanie. L’Europe. Troisième édition – ��-�� Septembre ����. Tome III: Section IIIA: Le Veda-Vedāṅga et l’Avesta entre oralité et écriture (ed. by J.E.M. Houben & J. Rotaru): ���-���. Bucarest: Bibliothèque de Bucarest. 190  From Fuzzy-Edged “Family-Veda” Ratnagar, Shereen. ����. “Agro-pastoralism and the Migrations of the Indo-Iranians.” India: Historical Beginnings and the Concept of the Aryan: Essays by Romila Thapar, Jonathan Mark Kenoyer, Madhav M. Deshpande, Shereen Ratnagar. New Delhi: National Book Trust. Renou, Louis. ����. “Les connexions entre le rituel et le grammaire en sanskrit.” Journal asiatique ���: ���-���. [Reprinted as: Renou ����: ���-���.] ——— . ����. Les Écoles Védiques et la Formation du Veda. Paris: Imprimerie Nationale. ——— . ����. Louis Renou: Choix d’Etudes Indiennes. Réunies par Nalini Balbir et Georges-Jean Pinault. Paris: Ecole Française d’Extrême-Orient. RPrāti = Rgveda-Prātiśākhya. Ed. and tr. by Max Müller: Rig-Veda-Pratisakhya, ˚das älteste ˚ Lehrbuch der vedischen Phonetik . Leipzig: F.A. Brockhaus. RV = Rg-Veda. ˚ ˚ RVāśv = Rg-Veda, Āśvalāyana recension: The Āśvalāyana-Saṁhitā of the Rgveda ˚With Padapāt ˚ ha, vols. �-�, ed. B.B. Chaubey. Delhi: IGNCA / D.K. Printworld. ˚ . Sachau, Edward C. ����. Alberuni’s India, � vol., London: Trübner [Reprint of the ���� edition: Delhi, S. Chand & Co., ����]. Scharfe, Hartmut. ����. Grammatical literature. A History of Indian Literature, Vol. �, fasc. �. Wiesbaden: Harrassowitz. ——— . ����. Education in Ancient India. Leiden: E.J. Brill. Smith, Frederick M. ����. “Exchange Rates in the Vedic Ritual: The Mūlyādhyāya- pariśis.t.a of Kātyāyana.” Wiener Zeitschrift für die Kunde Südasiens ��: ��-��. Staal, Frits. ����. The Fidelity of Oral Tradition and the Origins of Science. Amsterdam: North-Holland Publishing Company. Stutley, Margaret. ����. Ancient Indian Magic and Folklore. London: Routledge & Kegan Paul. Thieme, Paul. ����. « akhkhalı̄kŕ. tyā ». Zeitschrift für vergleichende Sprachforschung auf dem Gebiete der indogermanischen Sprachen ‘Kuhn’s Zeitschrift’, ��: ���. Thompson, George. ����. “The pursuit of hidden tracks in Vedic.” Indo-Iranian Journal ��: �-��. Verhulst, Pierre-François. ����. “Notice sur la loi que la population suit dans son accroissement.” Correspondance Mathématique et Physique de l’Observatoire de Bruxelles, ��: ���-���. Wackernagel, Jakob W. ����. Altindische Grammatik. Band II, �. Einleitung zur Wortlehre - Nominalkomposition. Göttingen: Vandenhoeck. 191 Jan E.M. Houben Witzel, Michael. ����. “On the localisation of Vedic texts and schools (Materials on Vedic śākhās, �).” In India and the ancient world. History, Trade and Culture before A.D. ���. P.H.L. Eggermont Jubilee Volume (ed. G. Pollet): ���-���. Leuven: Departement Orientalistiek. ——— . ����. “Tracing the Vedic dialects.” In Dialectes dans les littératures in� do-aryennes (ed. by C. Caillat): ��-���. Paris: Institut de Civilisation Indienne. ——— . ����a. “Early Indian history: Linguistic and textual parameters.” In The Indo-Aryans of Ancient South-Asia (ed. by G. Erdosy): ��-���. Berlin: de Gruyter. ——— . ����b. “Rgvedic history: poets, chieftains and polities.” The Indo-Aryans of ˚ Ancient South-Asia (ed. by G. Erdosy): ���-���. Berlin: de Gruyter. ——— . ����. “The Development of the Vedic Canon and its Schools: The Social and Political Milieu.” Inside the Texts, Beyond the Texts: New Approaches to the Study of te Vedas (ed. by M. Witzel): v-xix. Cambridge, Mass.: Dep. of Sanskrit and Indian Studies, Harvard University. ——— . ����. “Gandhāra and the formation of the Vedic and Zoroastrian canon.” In Travaux de Symposium International: Le Livre. La Roumanie. L’Europe. Troisième édition – ��-�� Sepembre ����. Tome III: Section IIIA: Le Veda-Vedāṅga et l’Avesta entre oralité et écriture (ed. by J.E.M. Houben & J. Rotaru): ���-���. Bucarest: Bibliothèque de Bucarest. 192