The vast quantity of ancient Indian material relating to myth, legend, and ritual metaphysics (SaMhitaas and BraahmaNas), cosmological and existential abstractions (AAraNyakas), and philosophical disputations (UpaniSads); to say nothing of the myriad ritual Suutra's (Shrauta and GRhya), disciplinary didactics (Vedaa^Ngas), and multitudes of orthodoxies (darshana's) have made comprehensive examination of even the most central terms and concepts impossible. Recent strides in both technology and methodology throughout the Indological disciplines have brought the field to the threshold of substantial breakthroughs in the understanding of how key ideas developed, changed, shifted in terminology, and interacted with social and geographical phenomena over the last 4000-5000 years of South Asian history.

    The present study is an effort to bring the fruits of the last 150 years--and especially the last two decades--of research on early Vedic materials to bear upon the question of the earliest discussions of existence and individual experience--that is, of the self--in the earliest period of Vedic literature. While the vocabulary has been either taken for granted due to its predominance in the later Vedic literauTre--as with aatmán, púruSa, and bráhman--or obscured by changes during the earliest periods--as in the case of tanuú and tmán--the context from which the influential philosophies of ancient and classical India arose has been paradoxically both taken for granted and largely ignored.

    The changing use of terms related to the notion of self provides a record from the earliest period of Vedic literature of different ways in which the self was identified in the early history of Vedic religion. The ideas of individual existence represented in the terminology were selectively fused together or discarded throughout the process in which the literature was composed and redacted. For instance, aatmán and púruSa are relatively uncommon terms until the later RV. They appear in a handful of largely unrelated contexts while tmán is quite common throughout this text. However, in the next period of literature in which the sacrificial texts of the Yajur Veda school predominate, tmán dissappears while aatmán and púruSa are repeatedly associated together with one another as components of a composite notion of self.

    By combining synchronic analysis of the occurrences of the terminology in a given passage with diachronic comparison of these occurrences between texts and over time this study shows that there were several developing--or even competing--ways of discussing existential presence in Vedic religion. The most direct means to illuminate these developing notions of the self with minimal presupposition is to comprehensively analyze each passage which includes one or more of the terms--over 15 words--chosen for examination in this study.

    The simplicity of the question underlying this dissertation belies the complexity involved in its answer: What was the liturgical and linguistic setting in Vedic religion from which the later speculations on the self developed? Though the terminology related to individual existence employed in both the later Vedic texts and the commentaries upon them was also used in the earliest known sources, a review of this early literature (c. 1500 b.c.e. for the RV) reveals that a greater variety of words was associated with individual existence than are found in the UpaniSads and philosophical commentaries.

    This dissertation, then, is an historical inquiry into the earliest collection of terms related to the self in the Vedic literature. It examines the Vedic origins of the concept of individuality, the distinction between living and non-living, the mental processes by which existence was suggested or explained, and the existential speculations regarding the relation between deity, individual, and environment. The material to be examined focuses on the Rg Veda followed by those texts compiled immediately after it--the Rg Veda Khilas, the Atharva Veda, as well as the Black and White Yajur Vedas--with attention to how these early ideas were treated in the BraahmaNas. These texts comprise the earliest speculations from which the majority of later darshanas chose aatmán, púruSa, and bráhman as their primary terms of existential reference and metaphysical discussions. The Early Vedic material, however, contains a wider array of terms with varying significations of self, presence, individuality, or life force. It is the examination of this wider array of terminology in the Rg Veda (RV) and its relation to the terms of later predominance in the subsequent Vedic literature that forms the central task of this study.

    The relative chronology of the early Vedic literature forms the central structure by which I will examine the Vedic material. An important aspect of this dissertation is the application of the most precise available sequence of the actual chronology of the Rg Veda and the texts that follow it. For the

most part, this has not been used by scholars for whom the traditional progression (Max Müller, 1968)--Veda's, BraahmaNa, AAraNyaka's and UpaniSads--was sufficient to support the foundations of Vedic research.1 Nonetheless, as early as the later 19th Century 2 scholars began dissecting the Vedic literature with attention to the complexity of internal chronology within each text. Thus the RV consists of several distinct chronological strata of organization, the oldest of which is MaNDalas II-VII, the so-called Family Books.3 As a result of recent research, there is a greater degree of detail available to scholars for examining the early development of key terminology than has, for the most part, been used in Vedic scholarship in the past decades. In addition, the rediscovery of internal chronology in the Vedic texts has been facilitated by the availability of additional texts enabling scholars to compare styles and periods of linguistic change, such as the Atharva Veda Paippalaada SaMhitaa, the KaaThaka SaMhitaa of the Black Yajur Veda, the Jaiminiiya SaMhitaa and BraahmaNa, and varous Shrauta Suutra's (Witzel, 1997: 289).

    The discussion of individuality in the RV is marked by changes in vocabulary between the early and later portions. In point of fact, terminology for the self in the RV is marked most prominently by tanuú and tmán which are substantially more common than aatmán and púruSa. By contrast, the earliest portions of the RV reflect only two instances of aatmán, and a handful of púruSa, while bráhman occurs with great frequency, though not in a context of metaphysical speculations regarding identity of the self with the cosmos. Later, throughout the BraahmaNa's and UpaniSads, to say nothing of the later commentaries and philosophical schools, there was little variance in the terminology for an individual's essence or presence.

    In most of the earliest UpaniSads--BRhadaaraNyaka, Chandogya, Taittiriiya, and Aitareya, for example--aatmán predominated in speculative debates of Yaajñavalkya, Janaka of Videha, Shvetaketu, Uddaalaka, etc. For the early RV, however, in the place of aatmán, tanuú and tmán appear as the words of choice, predominating in frequency of use. Beginning with the later Rg Veda and continuing through subsequent Vedic literature, tanuú changes in meaning and then, with tmán, falls out of use altogether. Tanuú''s predominance as an early term throughout the RV is also contrasted in the early literature of the BraahamaNa period. Occasions of tanuú in the Shatapatha BraahamaNa are less than one fifth in number as opposed to those of the earlier Taittiriiya BraahamaNa.4 Correspondingly,

aatmán overwhelmingly predominates in frequency of use over tanuú in the Shatapatha, while in the Taittiriiya BraahmaNa the relative frequency of aatmán is substantially less compared with tanuú. Both terms appear almost equal in number, though both are fewer in their respective number of occurrences as opposed to the ShB. This reflects a change in vocabulary and a corresponding change in the manner in which individual existence was conceived. Such changes have yet to be even recognized, let alone examined, prior to the present study.

    This dissertation integrates three tools of analysis as a model for precise study of Vedic material that correlates the data historically, linguistically, and--where possible--geographically. First, the most detailed data available for discerning the precise sequence of development in the Vedic literature is assembled here as a framework and research "control" for examining the terminology related to the self. This framework is presented in detail in Chapter 3. Second, with the aid of electronic research tools--Hypertext-Markup Language (HTML), electronic editions of Vedic texts, and electronic database sorting technologies (see below)--a level of detail and breadth of examination is made manageable which, until now, has required years of painstaking management of volumes of notes, texts, primary source editions, and references. Finally, the theoretical approach in the method employed here uses both the historical and electronic resources to integrate two vital research techniques: diachronic comparison over time within and between texts as to how a term is used, and precise synchronic semantic and linguistic analysis of each occasion of terms under study as it is used with those words forming the semantic field of its immediate setting.

    The terms to be examined include the primary group of words that are used in reference to the self--aatmán, tanuú, tmán, púruSa and bráhman.5 As part of the composite references to individual existence, the words referring to life--ásu, aayú, jiivá and praaNá--are considered as well as the words for corporeal or physical body--kraví, gaátra, déha, and sháriira. Finally, the words related to various mental processes or awareness--krátu, and those derived from the roots -cit, -dhii, -budh and -man--are also reviewed when they are found with the primary terms. I have chosen this group of words due to their prominence in later discussions of the self in the UpaniSads or--as with tanuú and tmán--due to their presence in semantic groups which are later used with aatmán or púruSa.

    This group of words represents pervasive or internal essence, life and

being-alive or vitality, comprehension and perception, reflexive self-reference (as most commonly demonstrated with svá and sváyam), individual identity or characteristics, and physical presence. These are the conceptual categories which most frequently figure into discussions of the self in the later Vedic and post-Vedic literature. In addition, there is a "catalogue of boons" from the Yajur Veda tradition in the Agnicaayana ("piling up of Agni" or building of the fire altar) which lists groups of benefits sought by the sacrificer. These include such things as cattle and worldly goods, ethical and moral uprightness, and so forth. The catalogue begins in each text with a groups of words related to the self, including the terminology chosen here (see Chapter 6 re. MS 2.11, KS 18.7, TS 4.7.1, and VS 18). Finally, my chosen terminology compares favorably when cross-referenced with the terminology chosen for examination in the previous studies which bear upon the self in Vedic literature (discussed below).

    The methodology for this study requires that a widely inclusive range of terms related to the self be examined across a range of genres and historical periods. Each term under study will, in turn, be analyzed synchronically with reference to words occurring in its "semantic field"6 (e.g. 1/4-verse, 1/2-verse, verse, hymn, etc.). Frequently the words adjacent to these terms for individual awareness are less ambiguous, affording a clear line of semantic analysis. The results of these synchronic, passage-by-passage analyses are then compared diachronically across different historical periods and text genres. Part of the synchronic analysis includes the identification of variant form--se.g., later development of the infinitive in -toH following the RV--that indicate particular historical or geographical information about the composition of the hymns.

    As I am not a linguist, the reader should know that I was led to incorporate this technical data as a result of my research into some of the more intractable passages which frequently included these exceptional forms. As the research progressed, a pattern of anomalies emerged which required some explanation. There was a remarkable frequency for linguistic anomalies in obscure passages dealing with the terminology for the self. I have followed this pattern wherever possible. The results suggest that, where two different dialect variations combine, the "fault lines" frequently intersect with the discussions of the self.

    It stands to reason that the symbolism which relates the microcosmic individual (usually the sacrificer)--in terms of body, mind, breath, social function, etc.--to the macrocosmic cycles of seasonal, celestial, and even

social order places a great deal of significance on the way in which the individual self is understood. For instance, the púruSa is given microcosmic and macrocosmic symbolism in RV 10.90, but the uses of aatmán suggest a less complex idea of vital or active essence. Later, the Black Yajur Veda ritual uses aatmán for the subtle associations of the sacrificer with Agni, while púruSa refers to the social aspect of dwelling in a house (Chapter 6). The problem, of course, lies in determining if and when aatmán and púruSa ever represented conceptions of the self from distinct peoples or dialects. Elizarenkova has suggested that púruSa might be a borrowing from another language (1995: 67), but this is still too vague for decisive conclusions.

    The field of inquiry wherein specific sociological, geographical, and historical information can be extrapolated from linguistic phenomena is still in its infancy. The significations of these anomalies awaits further research along the lines of Witzel's "Tracing the Vedic Dialects" (1989) and "The development of the Vedic Canon and its Schools" (1997) before it can be determined with precision whether competing doctrines of self account for the variations or whether they are all chance idiosyncrasies. For this study, I will simply note the patterns of linguistic anomalies where they are evident, though always with the awareness that there could be greater significations of differing religious perspectives on the self at work.

    The trajectory of the study and its results are organized according to the historical line of sequential development within and between each text. The dissertation is roughly divided into two categories. The first three chapters present the methodology, tools for analysis and the historical framework of the Early and Middle Vedic literature. Consistent with the way each passage is examined, the synchronic tools of analysis are presented first in Chapter 2. This includes a summary of scholarship for each term as it was used in the RV and a corresponding functional definition with which to begin the study of each word in its immediate semantic field. This will serve as the starting point from which subsequent use of the words in each passage can begin. The second half of this chapter outlines the synchronic tools of the analysis: polysemy, synonymy, distinctions between divine and human language, phonetic choices, and the linguistic anomalies that identify specific historical and geographical periods in the literature. The third chapter presents the diachronic material for the analysis of the synchronic results. A detailed presentation of the internal chronology of the RV and the later developments of the Middle Vedic period provide both

the framework for this study and a useful summary of the wide-ranging bibliography of such research for Vedic scholars. The remaining chapters address Early Vedic in two detailed parts: RV 2-7 and the later portions of the RVMaNDala's 1, 8, 9, and 10 and also the "appendices" or RV Khilaani. The final chapter tests these results as to how compatible they are when applied to the occasions of each term in Middle Vedic, including the Black Yajur Veda, White Yajur Veda, Atharva Veda, and several BraahmaNas.

    The detailed analysis of each occasion of the terminology examined in the RV is made possible by HTML "links" (programmed connections between each word in an electronic edition of the text that enables movement from occasion to occasion instantaneously with a simple mouse-click).7 For example, each occasion of the word púruSa has been searched and identified in a line-by-line reading of the RV according to text chronology (e.g., 2-7 are the oldest MaNDala's, then 1.51-1.191 and 8.1-8.65--excepting 8.49-59--etc., see Chapter 3). Each identified occasion of puruSa was then linked such that a mouse click upon it will take you to the next occurrence of puruSa according to the actual historical sequence of the text.

    In this way each term can be reviewed without interruption across the historical course of its use in the given text, and can be cross-referenced with citations of one text within another (e.g., RV mantras used in the ShB).8 In so doing, multiple terms across many non-sequential periods in a text, as with the RV, can be mapped and then analyzed with regard to meter, adjacent vocabulary, period of composition, etc., with the ease of a mouse click. Additionally, sub-groupings of detailed studies based upon these initial inquiries are then automatically maintained in historic format and additional links can be continuously built--and evolved--upon the original study. As in this example, once occasions of púruSa were first marked, then these were reviewed (i.e., each link followed in historical sequence through the text) with an eye to occasions where púruSa occurs with -dhii, -man, -budh, and -cit, reflecting mental presence or awareness, occasions with words for body, etc.

    Immediately the challenge to this study in terms of its theoretical methodology arises when the question is asked: Does this suggest that the meaning and use of the terms changed while a universal concept of self remained the same? Or, does it mean that the concept of self changed such that different words from the terminological pool were required to keep up

with the changes? The results of the research presented in Chapters 4-6 indicate that the terminology reflects a change both in the way it is applied to discussions of individual existence and also a change in the notion of self that is discussed over time and from text to text. One or both of these possibilities have been addressed, at least three times, in this century. It is the purpose of the methodology discussed below to reconcile the apparent antitheses between these possibilities. First, I will present the current approach as it contrasts with the three studies that are closest to it in scope and subject.

Previous Studies

    There is no lack of studies on the idea of self in Indian religion. Most of these are focused on the UpaniSads and the later darshana's. There are also broad studies that consult--to some degree--the earlier Vedic sources. These studies do not work with the primary sources nor do they address historical development within or between texts in any detail. In addition, the Black Yajur Veda texts are virtually ignored in all related studies. Citations are from standard translations, which are not sensitive to the issues addressed here. The authors are inclined to predetermined agendas in their analysis and choice of data.

    Typical of studies with a predilection for reductionism and simplification of Vedic concepts according to Western paradigms is Troy Organ's The Self in Indian Philosophy. Organ does present an introduction in which he outlines the importance for the West of the study of the self in other cultures. Aside from never setting out exactly what he means by "philosophy," "self-knowledge," and "the East," he also feels that the steady "decline" of Western civilization is directly attributable to what Aurobindo and Radhakrishnan cited as the deliberate ignorance of the self in Western thought (1964: 14). Whether this ignorance does in fact exist--ironically, he nonetheless cites William James' four-fold notion of self among other ("deliberately" ignorant?) Western studies of the self--he makes it clear that the purpose of his study is not so much to identify the notion of self in India as it is to extract one as a model that will fit the absence of one which, he perceives, persists in the West.

    Organ's approach, which seeks the self in some abstract, transplantable, or even universal concept, is a limiting factor in many of the studies discussed here. It leads Organ to suggest--without use of primary sources and only limited, non-critical translations--that a spirit-body dualism is

implicit in the Rg Veda: "These [Rgvedic] quotations indicate that the rishis believed man is a self which is other than the body" (1964: 28) He neglects the other SaMhitaas altogether, as well as the BraahmaNas and AAraNyakas, moving directly to the UpaniSads where he reads them according to Sha^Nkara, without the benefit of examining the texts themselves.

    In a more recent related study, N. Ross Reat has provided a sound, systematic analysis of early South Asian discussions of individual existence and psychology in The Origins of Indian Psychology. This study is one of the few that addresses a wide range of terms beyond those traditionally associated with the self--aatmán, tanuú, tmán, and púruSa--instead including jiivá, praaNá, ásu, and so on. His basic approach is based on the assumption of Heinrich Zimmer that all Hindu thought does not come solely from the Vedas but from both the Vedic and non-Vedic traditions. Unfortunately this prompts him to limit his text pool to the RV for early Vedic origins as it is the "purest," or least influenced by non-Vedic traditions ". . .since one of the major goals of the present study is to distinguish Vedic from non-Vedic characteristics in Indian psychology" (1990: 3). It prompts him to have "ignored to a large extent" (1990: 4) the other SaMhitaas, the BraahamaNas, and the AAraNyakas. The starting point for his investigation of the origins of Indian psychology is the monism of the UpaniSads, and he reviews the RV with the intention of finding the origins of this "monistic absolute" (1990: 9).

    Reat offers a useful discussion of a few RV passages, focusing primarily on building a composite notion of identity around the heart (hRd),9 cít/thought (1990: 99), -man/to think and mánas/mind (1990: 97ff.), tanuú/a "quasi-material," "personification" (1990: 63-64),10 and aayú/duration of life (1990: 84). He relies on translations for this analysis--Griffith and Muir--while nonetheless attending carefully to a substantial term pool. His summary of the functional roles of tanuú, shariira/physical body (1990: 69), ruupa/body or appearance (1990: 70-71), naama/name (1990: 74), the "Vital Faculties" jiiva/alive (1990: 81) , ásu/vitality (1990: 82), váyas/food (1990: 89), and praaNá/breath (1990: 91) are quite helpful as starting points for further investigation. Nonetheless, his ultimate objective is to analyze them according to a pre-determined idea of monism in the Rg Veda as the foundation for a conception of psychology in ancient India. He is not engaged in systematic analysis of the texts or terminology. This leads him to make broad generalizations concerning the Vedic terminology, suggesting that bráhman,

aatmán, and púruSa are all equally interchangeable words for the "unitary principle of the universe" (1990: 78). Yet he still feels confident enough to suggest that the terms "overlap considerably" and are "non-technical" in nature (1990: 63).

    Nonetheless, it is worth pointing out--in contradistinction to the other scholars cited--that Reat scrutinized the word tanuú fairly closely. Reat does not accept the view that tanuú means only "body." Based upon his analysis of "afterlife" in the Rg Veda, he suggests " . . . it should be clear that in Vedic thought the most essential element of the human being is his individual identity, which resides primarily in the quasi-material tanuu" (1990:63). It is "the identity link through which [the] carry-over into an afterlife was thought to be accomplished" (1990: 64). In fact, Reat feels that human existence is not so different from that of the divine: "Not only are humans like gods; they are intimately linked to them, and to the cosmos in general, in origin and essence" (1990:58). This link is facilitated by tanuú, a term to which Reat alone attends in detail.

    The significance of this attention is further underscored by a review of Indological literature as a whole (Dandhekar; 1973, 1978, 1985, 1986, 1993). There are a handful of isolated studies of tanuú with regard to tantu and the "warp and woof" on which the cosmos is woven. Otherwise, and particularly with regard to the studies of the self, the term remains largely taken for granted and unexamined apart from a recent internet discussion on the Indology listserve. Among other useful notes, the electronic mail discussion confirmed the paucity of such studies while underscoring the meaning of -tan and its sense of "to spread/stretch forth."11


    Unlike the other scholars discussed here, Reat offers citations drawn widely from throughout the RV (as opposed to later RV mantras in 1 and 10). Unfortunately, however, he does not show any awareness of historical development or relative chronology within the text. He has chosen to focus on the RV alone, and does not address any subsequent literature except the UpaniSads. This leaves his findings, which have little linguistic grounding, without any systematic or historical rigor. His work is not systematically traced through the texts according to their historical development or with regard to the changing terminology surrounding each word he examines. Because he does not work with primary sources, idiosyncratic uses of the terms are not accounted for. Of course, it should be noted that systematics of this kind are not his main focus. His predilection for demarcating an arbitrary, undefined set of "Vedic" ideas from "non-Vedic" bring to his

study a useful perspective, but it is otherwise somewhat ancillary to this dissertation. In addition, he never specifies what he means by "non-Vedic," though he considers the Atharva Veda (AV) quite prone to "non-Vedic" material (1990: 3-4).

    Apart from these admittedly partial instances, publications offering detailed linguistic and semantic analysis of the terminological contexts from which the broad themes relating to the self arise are largely non-existent.12 Some of the studies that do attend to the semantic and syntactic data in detail, such as Elizarenkova's Language and Style of the Vedic RSis, are indispensable tools for understanding the wide range of syntactic devices employed in the composition of the Rg Vedic vision. There are also isolated articles addressing particular hymns of the RV that also include some of the wider pool of terms. Many of these studies offer a specific view representing a particular school of interpretation, frequently that of Sha^Nkara's Vedaanta. Studies that address the historical question of development surrounding any one of the many terms used in relation to the self are scarce.

    There are only three studies that address any of this terminology in detail. Two of these are concerned with aatmán and one with púruSa. Of the three, excepting the earliest study--that of aatmán --by H. G. Naraharithe issue of the wider terminological pool is still left largely unaddressed. There is some analysis of usage across several texts in T. Sahota's study of púruSa, and a little in B. R. Sharma's survey of pre- UpaniSadic uses of aatmán. However, little attention, if any, is given to the distinctions of literary or ideological genre and the varying composition periods over time in each text.


    In 1944, H. G. Narahari published a detailed recapitulation of several previous articles (1942a, 1942b) in which he proceeds with the assumption that aatmán means "soul" in early Vedic texts, primarily the Rgveda. His book, AAtman in pre-UpaniSadic Vedic Literature, which grew out of his dissertation and other related articles offers his conclusions that aatmán was known to the Vedic R'Sis as "the spirit [which] is something . . . entirely different from the body" (1944, p. 15). He suggests that the aatmán-bráhman relationship was part of their speculations along with immortality, heaven, and traces of transmigration and kárma theories, and--interestingly--that UpaniSadic thought was not the result of a sociological upheaval in which the kSátriyas supplanted brahmín authority in the hierarchy. Instead, Narahari feels the role played by the kSátriyas was

"allegorical" to indicate the applicability of the refined doctrines to everyday problems. Against the idea that the wise kshátriyas represented any kind of revolution he suggests that there are as many ocasions when a king calls on a priest for wisdom as there are of the king correcting the priest (1944: 162ff).

    He arrives at most of his conclusions regarding the self by way of a detailed analysis of the occasions in the RV where aatmán is used. Apparently, this is primary word which he recognizes--to the exclusion, for instance, of tanuú--as representing the self in the Vedic literature. He is only somewhat attentive to the wider pool of words that surround aatmán. Ultimately, his work often seems more quantitative than analytical as his first chapter is entirely devoted to cataloging the various forms of the words he is reviewing. He includes aatmán, cít, ajóbhaaga, tmán, jiivá, sátya, mánas, and ásu. Narahari concludes this "conspectus" of terminology by suggesting that

The Rigvedic seers can thus be credited with the knowledge of the following: (1) that there is some "Spirit" or "Self" in man; (2) that it is different from the body and survives the destruction of it; (3) that it is eternal, neither born nor liable to destruction; (4) that it forms the 'essence' of the body and is its controller; (5) that it is the experiencer of the reward of man's actions i.e., Heaven or punishment after death; (6) that it is composed of the three qualities, Sat, Cit, and AAnanda (1944: 15-16).

He also suggests that it was the work of the UpaniSadic seers to simply consolidate and further systematize all this pre-existing doctrine.

    Narahari presents a good model for the present study with his multi-term analysis, but he remains focused upon locating the sense of aatmán as he perceives it to have been used later in the UpaniSads. One notices that his second chapter, after introducing a quantitative survey of the uses of aatmán in the RV, provides an extensive study of the later UpaniSadic applications of the word in relation to bráhman. Further, his citations in support of the "pre-existing" doctrine of aatmán (i.e., sát, cít, and aanandá) are almost exclusively from the first and last maNDala of the RV.

    A major limitation in Narahari's study lies in his inattentiveness to issues of historical precision in his analysis of the literature. This lack of attention takes two forms: failure to recognize the internal chronology of the RV, and mistaking later doctrinal views as appropriate guiding assumptions for his inquiry into the earlier texts. With respect to the first, he fails to see earlier and later strata of composition in the RV: all parts of the text are treated as part of a unitary, monolithic whole. Second, his inquiry is guided by a hindsight that seeks the origins of aatmán as it came to be

understood the later texts, largely according to Vedaanta. He sees aatmán as it was defined in the 19th-century UpaniSadic scholarship of Deussen13 and others as "the soul," - something distinct from the body.

    In his dissertation at Kyoto University in 1956, T. Sahota explored The Development of the Concept of PuruSa, offering a study very much like this one with regard to his analysis of terminology across multiple Vedic texts. Handwritten, with only one copy outside of Japan,14 it is in Japanese which I cannot read, but the romanized citations and references yield several observations. Sahota is attentive to the uses of púruSa throughout the early and later Vedic literature, though he seems somewhat concentrated upon the BraahamaNa period and later texts. A remarkable percentage of his RV discussion appears directed to hymns 10.71-72 (neither hymn contains púruSa, both contain tanuú), 10.81 (one occasion of tanuú, two of mánas, and none of púruSa), 10.121 (no occasions of púruSa, but aatmán, praaNá, and Prajaapati are included), and 1.162-164 (there are no occasions of púruSa in all of MaNDala I, but several occasions of aatmán, tanuú, and sháriira). It is not surprising, however, that Sahota's study reflects a shortage of hymns containing púruSa because, as noted in Chapters 4 and 5, there are only 23 occasions of púruSa in various forms, and 7 of these are in the PuruSa Suukta, RV 10.90. In addition, with the exception of RV 10.71, the hymns he chooses are those cited in the Shatapatha BraahmaNa (ShB).

    Additionally, this prevalence among Sahota's RV citations of mantra's used in the ShB indicates not only a concern for the later literature, but his predominant interest in the development of the sacrificial cosmology through the representation of the cosmic person--or puruSa. He also discusses puruSamedha/human sacrifice on several occasions as he moves deeper into his examination of the ritual texts in which púruSa comes to figure prominently in the sacrificial cosmology.

    Take, for example, his extensive use of RV 10.121, beginning with the second half of the 9th verse:

yásh caapásh candraá bRhatiír jajaána kásmai devaáya havíSaa vidhema ||

". . . (he) who birthed that shining great water, who is the god that we should worship with our oblations?" Sahota includes this question from the hymn to Ka/Prajaapati in his initial introduction of the puruSamedha citing ShB 13.6.1-2 (1956, I: 68). Following this discussion he begins an examination of Viraaj15 by way--it seems--of addressing his concern with

the multiple doctrines of sacrifice echoed in the literature. Citations for his other publications, also in Japanese, have been summarized by R. N. Dandekar and the summary shows a tendency in Sahota's work--including his dissertation--to examine doctrinal differences along caste lines (e.g.,"On the KSatriya origin of the UpaniSadic philosophy" [1978: 83]). Sahota continues to address various occasions of multi-gendered birth/begetting cycles in his dissertation as with Viraaj in BAAU 4.2.3, and with DakSa/Aditi by way of RV 10.72.4:

bhuúr jajña uttaanápado bhuvá aáshaa ajaayanta | áditer dákSo ajaayata dákSaad v áditiH pári ||

"The earth was birthed from the birthing womb,16 the regions were birthed from the atmosphere; from Aditi, DakSa was birthed, of DakSa Aditi also (was created) back again." In this case DakSa--who is named in RV 2.27.1 as a son of Aditi along with other deities known as the AAdityas (including Mitra, Aryaman, Bhaga, VaruNa, and AMsha in 2.27, their names and number varies elsewhere)17--has been born from Aditi, who in turn is born from him. Aditi herself is an abstract deity who is mostly known for her offspring, the AAditya's.18 Sahota follows the development of these trans-gender couplings through the BraahmaNa period citing Prajaapati's illegal coupling with USaa (AB 3.33), and that of Prajaapati and speech in PB 20.14.2.

    Sahota is concerned with the increasing prevalence of Prajaapati, and the corresponding development of monistic emphasis upon Vaac. Not surprisingly, he moves on to a discussion of RV 1.164 (1956, 1:75n) and 10.129 (1956, 1:85n). He then returns to 10.121, this time with the second verse and the occurrence of aatmadaá:

yá aatmadaá baladaá yásya víshva upaásate prashíSaM yásya devaáH | yásya chaayaámR'taM yásya mRtyúH kásmai devaáya havíSaa vidhema ||

Walter Maurer translates the passage as follows " - who is giver of breath,19 giver of strength; whose bidding all acknowledge, whose bidding the gods acknowledge; whose shadow is immortality, whose shadow is death. Who is the god we should worship with our oblation?" (1986: 268). Here with Ka exalted as the aatmadaá, or "self-giving," Sahota commences a lengthy discussion of the relation between aatmán and púruSa (1956, 1:88n).

    Sections II and III of the dissertation continue the discussion of

aatmán. The juxtaposition of both aatmán and púruSa occupies the remainder of the third section, with a persistent emphasis upon the occasions when both words are juxtaposed. Frequently, as here with AAA 3.2.4, his emphasis rests upon passages where one term is elevated or "marked" (e.g., with antar-, maha-, param-, adhi- or sarva- ), in contradistinction to the other:

sarveSaaM bhuutaanaamantarapuruSaH sa ma aatmeti vidyaat

"Let him know: the inner púruSa of all beings, that is my aatmán." The portion quoted by Sahota follows a lengthy exposition that "he/sa" is the one not heard, not thought, not seen, etc.; but who hears, thinks, sees, etc. This correlation of the inner puruSa of all beings is being related to aatmán in a passage on which A. B. Keith remarks (and is quoted in full by Sahota): "This is the most advanced point in the definition of the aatman arrived at (Keith, 1909: 254, n. 18) in the AAraNyaka" (1956, III: 66). As the dissertation moves to its conclusion, then, Sahota begins a discussion of "puruSa-vid" and "aatma-vid" in which Viraaj figures prominently (1956, III: 102-105). Opportunities to directly compare aatmán and púruSa in the same passage are rare--only twice in the RV at 10.97.4c and 8c, which Sahota does not address at any point in his otherwise apparently thorough study!--and these are addressed in Chapter 5. Sahota's work remains in need of a proper translation as even a survey of its contents in Sanskrit, German, and French suggest that he is covering ground that has not been considered, apart from the present study.

    Based on a survey of the database and of Vishva Bandhu's Vaidika-Padaanukrama-KoSaH (1973) it is apparent that neither aatma-vid nor puruSa-vid were terms of parlance in either the SaMhitaa's or BraahmaNa's. PuruSa-vid is found once in Maitri U. 6.33, discussing the three breaths and the sacrificial fire. AAtma-vid is more prevalent in the later literature of the UpaniSads, but has only one occurrence earlier in ShB which includes the following litany of kinds of knowledge:

brahmavitsá lokavitsá devavitsá vedavitsá yajñavitsá bhuutavitsá aatmavitsá sarvavid This passage occurs in a debate where Yaajñavalkya

questions Patañcala Kaapya about whether he knows the thread/suutra by which the world is held together (cf. BAAU 3.7.1, also 3.3.1). In the case of ShB, the question is asked about whether Kaapya also knows the antaryaamín, the inner course or controller of all things. The knower would then possess

all the knowledge enumerated above. It is worth noting that one of the few occurrences of either term occurs in a "knowledge catalogue" of sorts, which does not list puruSa-vid.

    It appears that Sahota has either adopted both terms out of context as a means to explain the distinction in doctrines--that of aatmán vs. that of púruSa--or as a foil that he uses to argue for a homogenous sacrificial doctrine regarding individual presence in the ritual cosmos. It is more likely that Sahota argues for the former. This is suggested as well by his citations from Winternitz (1972: 167) and Oldenberg (1888) which emphasize the panoply of ideas from which the Vedic literature was codified (1956: I, 86ff).

    The suggestion that appears to be a central element of Sahota's thesis--that puruSa-vid represents a distinct school of thought from aatma-vidis quite viable. This is discussed in more detail in Chapter 5, but there are other scholars who have considered that there were distinctly different groups of priests whose ideas were incorporated into the ongoing Vedic tradition. The research supports this in several ways.

    By way of introduction, consider the prominence of púruSa in the later schools such as SaaMkhya and Yoga, also in the Shvetaashvatara UpaniSad. However, when comparing aatmán and púruSa in the RV, aatmán is used twice as frequently as púruSa. Yet the semantic fields surrounding aatmán do not in any way represent a subtle or abstracted meaning such as became prominent in the later literature. However, the most frequent use of púruSa, found in RV 10.90, displays an extremely advanced micro-macrocosmic symbology already in place. It is also curious to note that the entire 9th MaNDala does not contain a single occasion of púruSa (in addition, the Saamaveda hymns that do are not repetitions of any other part of the RV hymns which contain púruSa). Also, there are only eight occasions of púruSa in the Family Books of the RV, and they contain no acknowledgement of púruSa as anything more than a human--e.g., manuSya, puMs-, etc. As noted above, Elizarenkova has suggested that púruSa, for instance, may well represent borrowing from another language (1995: 67). Consider also the suggestion by Frits Staal that (following Kosambi, 1950), "Vedic brahmins were to a large extent recruited from the priest class of the conquered pre-Aryan population" (1983, I: 138). The resolution of the question insofar as Sahota has or has not addressed it, of course, awaits the knowledgeable reading of a scholar conversant with Japanese.20

    In a similar vein, Baldev Raj Sharma offers a study quite like Narahari's, The Concept of AAtman in the Principal UpaniSads (in the Perspective of SaMhitaas, BraahmaNas, AAraNyakas, and Indian Philosophical Systems). While he is careful to outline an array of possible terms related to the self, he does not follow through on their analysis. He is familiar with Narahari's work, citing his quantitative results, and listing him with Edgerton (1916: 1977ff.), Bloomfield (1930: 220ff.), Ranade (1921: 3f.), Radhakrishnan (1956: 72), and Keith (1925: 493f.) as an authority who controverts "the majority of scholars [who] without adequate appraisal of the Pre-UpaniSadic Vedic texts came to the much doubtful conclusion that the philosophy in the real sense started in India with the UpaniSads" (1972: 29). As the analysis below in Chapters 4-6 indicates, he is quite right in this observation. Simply because some of the later philosophical vocabulary--aatmán and púruSa--is largely absent in the RV, this does not mean that there is no philosophy as well. As demonstrated in Chapter 4, other terminology--specifically tanuú and tmán--serves to portray a complex interrelation of the divine and human realms. This is not Sharma's conclusion, however. Sharma chooses to equate early terminology with aatmán and, in so doing, affirms that the doctrine of the UpaniSads is argued consistently in every text of the Veda's.

    While correcting the idea that the Vedic texts were lacking philosophy is surely warranted, throughout his study Sharma seeks the aatmán as it became known in the later texts, or, as he translates it, "Ultimate Reality." He seeks a "clear and comprehensive meaning of the concept of AAtman in the UpaniSads scientifically through a comparative and historical study of the various views on the subject" (1972: iii). While there were originally many terms for this "Ultimate Reality," they came to be "dropped (due to semantic changes), or replaced by some new terms and names" until the words aatmán and bráhman "became standard for the expression of Ultimate Reality in the UpaniSads" (1972:8-9). This predilection to find one later meaning--Ultimate Reality--in all words related to individuality in the Vedic texts renders Sharma's study problematic for the objective historian, apart from the other methodological concerns described below. He asserts:

In the SaMhitaas, like the word Brahman, which is equated with AAtman, the words: puruSa, haMsa, suparNa, ajobhaaga, praaNa, jiiva, satya, Vishvakarman, BRhaspati, Prajaapati and hiraNyagarbha, and so on, also denote the sense of AAtman. . . . While UpaniSads breathlessly used the terms AAtman and Brahman to denote the Ultimate Principle, the SaMhitaas had their own terms to express the same thought " (1972:14-15, [emph. mine]).

He subsequently suggests that these terms "stand as equivalent to AAman in the Vedas" (1972:15; cf. also a similar tendency in Reat, above).

    However, Sharma does a masterful job of drawing together references for the variety of words listed as synonymous to aatmán in Nirukta 14.10,21 considered by some as a later addition (1972: 23). His study surveys a relatively large portion of the Vedic literature. Nonetheless, he relies primarily upon hymns from the first and tenth maNDalas of the Rg Veda. From the other SaMhitaas he chooses a handful of hymns based on his perception of their overall theme regarding "Ultimate Reality." He does not deal with any other terms in detail. In addition, it is apparent that he is not aware of historical sequence of composition--certainly within the RV--and has only a general sense of relative sequence for the early Vedic texts of the other branches.


    It would appear from this survey of previous scholarship that the paucity of comprehensive analytical studies of this question of the self in the early Vedas has arisen more from methodological challenges than from inattention to the importance of the question. It is more than feasible to examine one term in isolation but the choice of terms examined thus far reflect the perspective of the later UpaniSadic traditions. It would almost seem as though for scholars of the Veda's a notion of self had not "appeared" until the terms of later predominance--aatmán and puruSa--came into popular parlance.22 More importantly, any trajectory of change or development is lost in every one of these studies due to the lack of awareness--or inattentiveness--to relative chronology within the texts, particularly the RV.

    In the case of each study there were elements of the current approach, but inevitably all were oblivious to successive periods of composition within the RV. In addition, while all were attentive to other related words, these were not studied systematically. Excepting Sahota, all studies sought a single axiomatic meaning for the term under study. For Sharma and Narahari, this was a later doctrine of aatmán as understood in the UpaniSads. It is also interesting that neither Sharma nor Narahari even mentions tanuú. Sahota does so only once, directly after his conclusion regarding aatmadaá, púruSa and aatma-vid (1956, III: 106).

    One study that stands as a transition to the method employed in this dissertation from those above is Jan Gonda's Notes on Brahman (1950). Gonda is not particularly concerned with the notion of the self, and his

study of bráhman is directed more to its early uses and etymology as opposed to the later, more specialized, applications. In essence, he sees the term as derived from -bRh, to increase, expand, promote (1950: 69). He disagrees with Charpentier that bráhman derives from the terms for grass on the sacrificial vedi, barhis (1950: 5, 70-72), and with Hertel's assumption that it is fire (1950: 4). While Gonda says his purpose is not primarily etymological, he does devote a fair amount of time to etymology (1950: 39). He rejects Renou's conclusion (1949: 7) that it is an enigma, owing to Renou's proposed meaning that bráhman is power or expansion (1950: 57-58). He observes that it belongs to a type of Indo-European terminology that blurs the distinction between "nomen actionis and nomen rei," and thus that the distinction between action and object was not always clear (1950: 72). In this connection the enigmatic nature of power and expansion associated with bráhman is not unlike that of extension and expansion in the root -tan. In both cases, it is not clear where the "thing" (power, or expansion) itself ends and where the action entailed by it (growth, extension) begins. It is clear that definitions presented in isolation leave as many questions about a concept denoted by a term as they do answer. Accordingly, it becomes necessary to determine the function of these terms in relation to each other.

    Gonda observes, "Thus, one of the outstanding characteristics of Ancient Indian thought was a persistent adherence to the quest for knowledge of the mutual connection of all that has a name (whether it is, to our way of thinking, substance or attribute, spiritual or material, animate or inanimate, abstract or concrete), and is, consequently, a reality, a power, -- for knowledge of the inherent causality of things" (1950: 9). This attentiveness to connectivity stems from Gonda's larger methodological approach that significantly influences a major component of the approach in this study. He looks for meanings of words based on the associated ideas in each context--or "semantic field"--where the word, he calls this a key word, occurs (1962: 243). His article on the study of Indian religious terminology actually redresses some of the apparent contradictions in the bráhman study by way of clarifying the nature of his approach. He hopes to find ". . . how either traditionally or in a definite period, the Indians themselves thought about the basic, central or 'original' sense of a 'key word'" (1962: 269).

    Gonda cites texts throughout the Vedic corpus in his analysis without attending to the historical or genre distinctions between them. He goes so far as to say ". . . I criticized the main views upheld by my predecessors,

emphasizing the weakness of evolutionistic constructions and the difficulty of arranging the senses of ancient Vedic terms of outstanding importance, like bráhman, in such a manner that a definite historical development may be read off from the very arrangement" (1962: 267-268, cf. 1950: 4). Thus we are left with a compendium of comprehensive citations without any historical framework in which to understand one in relation to another.

    The strengths and weaknesses of these studies have been carefully considered in designing the content and method of this dissertation. There were a multitude of terms that were variously employed in the early discussions of individual presence and existence. It is also apparent that this term pool 'shrank' with the passage of time--or coalesced into the more prevalent aatmán, bráhman, púruSa, and, to a lesser or different extent, jiivá and praaNá groups. Beyond this, we have no systematic map of the pool of terms related to the self, no accurate or even vaguely specific understanding of historical development, and no analysis of the terms in context--apart from later doctrinal or thematic assumptions.

    Thus we are left with three issues to be resolved in the methodology as suggested by the studies listed above. First, the pool of terms chosen for study must reflect as broadly as possible the range of expressions related to individual existence and awareness in the early texts. This overcomes the problem of ignoring complex changes in the vocabulary and reduction of the findings to a single doctrine due to a myopic focus on a single word. Second, the terminology chosen for analysis must first be studied synchronically within the immediate semantic field of each occurrence. This prevents the interpretation of later developments (such as discussions in the UpaniSads or schools of Vedaanta) of the word--and the passages in which it is found--from influencing the analysis. Finally, the results of the synchronic analysis must be considered diachronically according to the historical development of the various texts. This creates an empirical, historical, and objective--rather than doctrinal--tool for comparing one occasion with another. In this way the results of the study are systematized along temporal rather than philosophical or doctrinal lines.

Methodological Challenges


    The first methodological challenge is the need to focus upon more than one or even a few words. Otherwise the study cannot adequately reflect the full context of expressions used with regard to individual presence in early and later Vedic literature. In place of continuing the ineffective

diachronic analyses of a single term--with or without the hindsight of a later philosophy or darshana--it is necessary to encompass the largest feasible body of words related to the self across the earliest texts and emerging literary genres of the Vedic period. As mentioned above, the primary pool of words chosen is ásu, aatmán, aayú, krátu, tanuú, tmán, púruSa, praaNá, and bráhman. I have added several observations about ahám and svayám to this primary base of words. To these are added a secondary tier of related conceptual term pools: those related to processes of thought derived from the roots, -man, -cit, -budh, and -dhii; and those related to physical or corporeal existence kraví, gaátra, déha, ruupá, and sháriira, largely in comparison with tanuú that later comes to refer to corporeal presence. To avoid the imposition of external or historically posterior perspectives, these terms under study--key words--must be examined within each context--semantic field--in which they occur.

    Secondly, there must be a timeline along which to organize the otherwise synchronous results of the analysis of key words and semantic fields. Recent studies have clarified the last 120 years of scholarship concerning the relative sequence of composition within Vedic texts whose contents until now have been treated homogeneously. These are addressed in detail in Chapter 3. As the fundamental "control" for this research, it has been attended to at length. Tracing the developing pool of terms related to the self in Vedic times requires a clear line of historical sequence in the development of the texts themselves. As mentioned above, for instance, periods of early and later composition in the RV have been known for over a century, but have only recently been consciously applied in Vedic Scholarship (e.g., Witzel, 1995a; cf. Chapter 3).


    This necessarily generates the third methodological problem, that of feasibility, a problem that has made studies of this kind quite cumbersome. Even if one were to study the full group of terms, doing so across the multiple layers of order in which a text like the Rg Veda was composed is a daunting prospect. This becomes even more problematic if several texts are under scrutiny across a wider period. Thus it has been necessary to use the technological assistance of electronic texts and hypertext links (HTML) between words to facilitate a working model of the available record of historical development for all eighteen terms for the RV and the ShB. For other texts not yet in electronic form, the VaidikapadaanukramakoSaH provides a ready resource of citations.

    Each of these three methodological problems have been addressed in

the design and implementation of this dissertation. The methodology combines the latest historical and linguistic evidence with regard to the sequential development of each text with a systematic analysis of the key terms under study and those that are commonly associated with them. The function of this methodology for addressing each will be discussed in turn with working HTML examples where appropriate.

Terminology Related to the Self and its Analysis

    Both the model and the terminology--"semantic fields" and "key words"--for the synchronic portion of the methodology are drawn from the work of Jan Gonda. Gonda, in a discussion of methodology suggests the need for addressing

. . . the distance in time, space, and cultural environment between Vedic mankind and the most modern specialists; the incompleteness of our sources; the reinterpretations suggested by the traditional views of the Indians; the prejudices and limitations of modern scholarship itself, which has often been guided by the tenets of contemporaneous philosophy, by the religious conviction of the research workers, or by the political systems of their own countries (1962: 244).

He affirms the importance of terminological studies because "our knowledge of, and insight into, Vedic religion largely depend on a correct understanding of a considerable number of Indian words and phrases" (1962: 243). To effect this "correct understanding," Gonda prescribes not only thorough philological and historical knowledge of the contexts and situations in which key terms under study occur, coupled with an understanding of the phenomenology of religion, but also "a readiness systematically to investigate the 'semantic fields' to which the term belongs and the cultural system to which it is related" (1962: 246).

    Nearly twenty years later, Gonda again reiterated the specific limitation of traditional lexical works (e.g., Graßmann) in which the same Western word is often applied as the first meaning to two otherwise different words, such as medhaa and maniiSaa both rendered as "wisdom" in the sense of "sound judgment" (1980-81: 3) In a further statement of his "semantic field" methodology, he suggests:

Instead of pursuing largely atomistic and often pseudo-historical methods we had better realize that in a given period and milieu words are used in synchronous systems as networks held together by differences, oppositions and distinctive values; that groups of words constitute 'lexical or semantic fields' intermediate between the individual words on

the one hand and the whole vocabulary of a given period and milieu on the other; that within such a 'field' the semantic areas of the individual units reciprocally limit one another. That is to say, we should first and foremost study the meaning of words synchronistically, regarding them as forming aggregations or associations of units between which there exist relations and connexions [sic] and attempt to understand these connexions, that is, the differences in use and meaning of the separate units, as well as the extent to which they are semi-synonymous, that is interchangeable (1980-81: 5-6).

Here, and in the article previously cited, Gonda owes much of his terminology to S. Ullman, whose work on semantics examines the historical development of semantic analysis considering the problems of cross-cultural and cross-linguistic studies. Ullman suggests such a synchronic approach to counter-balance the genetic and other fallacies inherent in diachronic studies (1951: 151f.) such as those listed above involving the influence of hindsight or later philosophical/doctrinal developments.

    In "Some notes on the translation of Indian religious terminology," Gonda cites how Vedic aMshu, which literally means "filament" or "filament of the Soma," is sometimes used by way of metonymy to mean the soma-juice. Little insight into Vedic thought or Weltanschauung is gained, he asserts, by simply labeling one instance of aMshu or another a case of metonymy. The question is "what made the Vedic poets use this word in what would appear to us to be 'two senses'" (1962: 247). In the present work, comparison between semantic fields, where the key word remains identical, will be used in conjunction with contextual analysis of the ritual, its component elements, the myth illustrated (if any), and related passages in other texts to illuminate the developing connections of concepts associated with a key word.

    Gonda's apparent unwillingness to take into account the historical developments when he deals with texts from widely differing temporal genres produces problems almost as large as those he seeks to overcome. His inattention to historical development proves counter-productive to the very aims he has set forth. As noted earlier, availability of additional manuscripts of previously unknown texts has enabled recent studies to have established a more precise sequence of development than--perhaps--was clear when Gonda articulated his method. However, Oldenberg's work was

done as early as 1888 and must surely have been known to Gonda. It is possible, however, that Gonda's unwillingness to reduce complex terminology to a singular "original" meaning might have accounted for his avoidance of chronological sequence in most of his work. The issue as to why other scholars have also remained unaware of this information is not clear.23

    Thus, in the analysis of AAA 2.6.1, Gonda makes no effort to distinguish the possible doctrinal and historical differences between the many texts used to explain the passage, and draws from the Rg Veda, Atharva Veda, Shatapatha BraahmaNa, and various UpaniSads indiscriminately as to their relative sequence or genre differences. Working in this way, Gonda blurs important historical developments by failing to consider the Vedas' own internal linguistic record of change and sequence. Perhaps Vedic kaala did develop along cyclical lines, but that did not eradicate the awareness among R'Si's of relative sequence and its function for identifying particular phases of doctrine or practice. However much a monolith the Veda might well have been considered, for the historian of religions this does not obviate the necessary distinctions of shaakhaa, relative chronology, and genre within it. This limitation in Gonda's work has not gone unnoticed by other scholars reviewing his research such as Stanley Insler (1993: 596-597), Richard W. Lariviere (1987: 837), and Joel Brereton (1988: 336-337; 1990: 369-370).

Timeline of Vedic Development


    It is important to maintain historical perspective for categorizing and systematizing data without also interpreting the chronological sequence as a causative factor in the development of the terminology for the self. Without this proviso the same errors as in the studies above, arising from hindsight, will be perpetuated and reiterated. It becomes too easy to review a text with later developments in mind. Instead, understanding that a given text has different chronological periods in its development gives diachronic structure and distinguishes difference in time within one genre when combined with detailed synchronic analysis. Synchronic analysis alone, however, can provide only a limited perspective on the concept under study. Gonda, who otherwise questions and even disregards historical categories in his research, notes that, "A thorough understanding of the literary peculiarity and significance of the Veda will however require supplementation by historical methods (1975: 59)."

    As has been apparent throughout the preceding discussion, applications of historical methodology can easily fall prey to a tendency to use history not as a perspective of inquiry, but as an axiomatic datum of explanation. Reading Vedic development of ideas with an assumed telos of historical causality was evident in the discussion of prior studies. To the credit of these authors, the refinement in scholarship as to the date and relative chronology of Vedic literature--even within a single text--has been a recent development. Implementation of this recent research forms the fundamental organizational premise of the current study.

    The idea that there is a sequence of broad text genres is hardly new to Indology. In point of fact, the idea that internal sections of a single text, or portion of a text, can each represent distinct ideological or historical phases of composition is not at all new. Besides the work of Keith (1909, 1914, 1920), Bergaigne (1878), Oldenberg (1888), and others on the internal chronology of the texts they either edited or studied, W. Norman Brown (1968) has offered an analysis and translation of RV 1.164 that seeks to understand its contents by identifying segments as separate additions, or periods of independent composition. He suggested viewing the contents of the hymn as consisting of independent units representing various periods. More representative of systematic study that yields earlier and later segments within the same text was the work of R.C. Hazra on the PuraaNa's (1975). In order to explain the variations in content and style within a single PuraaNa he suggests that it represents a composition of several periods, often separated by centuries.

    Careful consideration of known historical periods of composition provide a fundamental framework in which to place analysis. However, history cannot, at the same time, be the rule for interpreting the data without begging the question under study. The sheer volume of material written about the Vedic period necessitates this distinction in the methodology for this study. The material of the last century of scholarship concerning the chronology of the various segments of Vedic literature is provided in Chapter 3. As the specific dates for each text and segment still remain in much dispute, the more readily acceptable data of sequence will receive greater emphasis. To date, the most concise summary of all the relevant materials is that of Witzel and it is upon his conclusions, therefore, that the present study frequently bases its timeline.

    Witzel has used comparative analysis of Vedic linguistic phenomena to posit movements and change in three major Vedic population centers:

KurukSetra, Pañcaala, and Kosala-Videha (1989, also 1997). This careful linguistic work has opened the door to the reality of competing ideas, grammars, and their peoples (cf. Kenoyer, 1991: 332). He cautions on several occasions as to the importance of considering the literature within an accurate historical framework. Based upon the linguistic evidence, the relative chronology for the RV can be more precisely divided into groups of hymns comprising general periods as well as specific hymns that are later additions (1995b: 309).

    There are distinct strata of composition periods that correspond with linguistic and stylistic features. Witzel, following Mylius (1970: 423, also Narten, 1968: 115, n. 13), distinguishes between three broad layers of texts: Early Vedic (the RV), Middle (YV, BraahmaNas, and UpaniSads) and Late Vedic (suutras) (1995a: 97). In a more detailed chronological subdivision of the Vedic literature, he places the Rgveda alone as the sole text in the oldest category, Early Vedic. Next comes Mantra language, that of the AV, SV, and YV; followed by SaMhitaa-prose that is distinct from the prose of the previous category, distinguished in its content as expository, BraahamaNa-style discussions, in the MS, KS, KpS, and TS. BraahamaNa-prose includes two divisions of earlier and later: the older UpaniSads--BAAU, ChU, JUB, late BraahmaNa's, (GB), and the earliest of the Shrauta Suutras form this group. Finally in "Suura Language," the balance of the ShS and the GRhya Suutras, as well as UpaniSads of later origin like KaTha, Maitri, and Prashna UpaniSads are found. Only after this period do we find the clear emergence of PaaNini's local bhaaSa, Epic speech, and Classical Sanskrit. This will be of use below as Witzel notes that these three later forms are linked with the ShB and the AB (1995a: 96-97).

    I have found that the study of semantic changes can add a corollary to this kind of linguistic research. For instance, an hypothesis as to the later date of RV 3--later than all the Family Books--has been correlated by the distinct changes I have demonstrated in the uses of tanuú (see, for instance, summary in Chapter 3). As will be seen below, the HTML database differentiates each section of the RV as well as the relative chronological placement of the ShB, to be correlated with texts from later periods. Changes along that timeline in the vocabulary for the self can then be assessed and drawn upon to further refine and explore the religious developments of the period. This is made possible, in turn, by the wealth of stylistic analysis of the Vedic period as in the work of Elizarenkova (1995: 107), Gonda (1962: 243), and Renou (1938: 153ff.). who have pioneered exploration of the principles for linguistic analysis for the understanding of Early Vedic religion. With the exception of Renou, however, these authors have, at times, overlooked the historical develop

ment within one particular text (RV) and generalized across the entire text as though its content and composition represent a single period and perspective.24

    It is important to balance the variety of datalinguistic, historical, and philosophicalin such a way as to shed light on the actual conceptions of the self as presented in the Vedic terminology. Wilhelm Halbfass, in Tradition and Reflection, observes "In some central instances, the resolution of technical problems, and the attention to minute philological details, are indispensable in order to approach the broader issues. Philology and philosophical reflection cannot be separated in such cases" (1992: viii). In his recent study, On Being and What There Is: Classical VaisheSika and the History of Indian Ontology, Halbfass emphasizes the importance of linguistic analysis for the understanding of key concepts (in this case, "being"), citing John Stuart Mill, the Sinologist Arthur C. Graham, Willem Dilthey, and W. V. Quine among others (1992: 9ff). Quine, for instance, echoes Gonda's semantic field methodology calling for analysis of a semantic framework which discusses concepts such as being and existential participation (1961: 1f.).

    Thus, in this dissertation, the careful synchronic analysis of each key word in its semantic field, including lexical and linguistic considerations, will be placed according to the relative chronology of the Vedic corpus. This means not only attending to which text came before which, but which segment within each text precedes or follows another--either within or between texts. This is afforded by mapping data from the texts, by means of line-by-line analysis of the terminology in electronic format, with Hypertext Mark-up Language (HTML).

History of the Texts and Making the Links


    The combined impact of the issues and problems listed above has likely proven to be the primary reason why the current study will be the first of its kind. Additionally, of course, the advent of electronic text (e-text) manuscripts of primary sources, the World Wide Web (WWW), Hyptertext Mark-up Language (HTML), and the PowerPC are assisting in the resolution of many of these obstacles. In the present study, the results of historical analysis of internal and external evidence for establishing chronological date and sequence for Vedic literature, coupled with linguistic analysis of the relevant passages, will provide both the innovation and organization necessary for a detailed assessment of the Vedic conceptions of the self. These conceptions, derived

from semantic and syntactic analysis of the usage of key words related to individual existence, will provide a framework for sociological investigation of the changing currents of Vedic culture.

    The use of technology cannot replace detailed analysis. A "search and find" mechanism with an e-text--even if there were one that worked with the complexity of the Vedic alphabet and accent--cannot reliably identify the many possible variations in nominal declension or compounds, to say nothing of the intricacies of verbal inflection. Accordingly, nothing can replace the time-honored approach of detailed, line-by-line reading and identification of terms for study and analysis. The real problem comes when trying to extract and synthesize this data. A scholar is faced with cumbersome stacks of volumes that are hard to manage when the study embraces thousands of citations that are, themselves, a-synchronous as to the order of the pages in a text versus the historical sequence of composition within the text. It becomes logistically impossible to adhere to the actual historical development within a text.

    Indological lore tells the tale of Jan Gonda's study wherein one whole wall was filled with small boxes. As his scrutiny of a text yielded instances of one word or another, he would make an index card with appropriate notes, and place it in a box labeled with that word or concept. When a box was filled, he took the notes and wrote a book or article. Historical sequence is not easily facilitated by this method, as noted repeatedly above. With electronic texts, the boxes are replaced by historically sequenced HTML links. As new topics are considered, the original links can be instantaneously followed and annotated or augmented to take this new term or concept into consideration. The re-examination will still follow historical progression for the texts. With the analogue, or "paper" method, each note and box's collection represents a clutter of notes that must be re-sorted, a range of texts that must be re-assembled and examined, and a database which only has meaning--and to which access is effectively limited--for one scholar. It is idiosyncratic rather than empirical and systematic.

    For instance the RV, which has several periods of composition that are not reflected in its composite arrangement, makes detailed study of any concept or term--let alone several as in the present study--altogether impossible or prohibitively time-consuming. Not only would one have to read and make notes from the text, but this collection of notes would then have to be sequenced according to historical timelines, then be physically

accessed from volume-to-volume and note-to-note. This procedure is time-consuming for a scholar to manage and does not facilitate the ease of access needed for the scholar to have an adequate perspective to analyze the large quantity of data identified. The paucity of studies similar to this dissertation that consider even two words over a moderately broad historical range of materials is mute testimony to these logistical problems.

    The technology of electronic texts and HTML is not some convenient shortcut, it isas technological innovations have always beena tool with which new questions can be addressed for the first time. The most traditional research procedures were still employed at each step: line-by-line reading of the text, comparison between related genres, implementation of systematic criteria of synchronic linguistic and philological analysis, and carefully marking of these steps so each point of the process can be retraced for diachronic analysis. What is important about the way this dissertation employs the technology is not only its unique presence in the world-wide revolution of information access that the Internet and HTML represents. Instead, it is the first time that the traditional, proven methodologies of historical inquiry, linguistic and philological analysis, and primary source research are woven together in a systematic framework that is instantaneously accessible to any scholar with a computer and telephone anywhere in the world. The speed and precision of HTML links enable the research to be empirically verified and built upon systematically.

    The arrival of hyptertext and e-text databases simplifies large-scale studies, and offers a convenient, efficient, and rapid means of accessing the data such as that chosen for study. It can be cross-linked with related terms being studied within one text or throughout several. The concept, at least in a technological sense, originated in 1945, with the article "As We May Think," by Vannevar Bush. Bush was hypothesizing a "memex" as a solution to the "growing mountain of research," confronting and even overwhelming scholars. The memex would be a "sort of mechanized private file and library" (1945: 102). Its function would be to store, link and retrieve information. Espen J. Aarseth remarks, "His user is clearly modeled on the traditional academic author, who can carry out his critical comparisons and annotations of sources with the same serene distance as before, only much more efficiently" (1994: 68). The word "hypertext" was not coined until 1965.

    The title of Bush's article is somewhat ironic in connection with the creation in this dissertation of a "memex" for Vedic data. While "As We

May Think" is considered the "modern" origin of the concept (1994: 68), I would argue that--in Vedic terms--the title could well be "As They Did Think." Certainly it is a model of how modern thought is being conceived in the physical and neurological sciences. Theories of associative thinking and memory are common. In the present study, I am suggesting that this is not only a modern but a human phenomenon, and that this emulates how the Vedic R'Si constructed images of the universe and human presence within it. The development of the canon, and its existence and transmission in the form of an oral tradition, provide a dynamic database to the ya evam veda/"one who knows" from which a multitude of thematic, metrical, and semantic links could be drawn for ritual and meditative purposes. There are instances of this in which the BraahmaNas directly quote the RV as an invocation in a ritual, or the justification of mantra choice according to the content of a ritual or significance of a given word, and--especially as evidencing "developmental" thought--the brahmodya. At any given moment, the R'Sis had at their mental disposal an intricately interlinked database in their mind from which they could draw in expounding on a topic. With computer technology, I have built an electronic model of that canon to systematically and comprehensively trace developments within it.

    This is also consistent with the manner in which the mantras became composed according to the analysis of Elizarenkova. She describes the importance of multivalence in vocabulary, particularly in the form of polysemy. The R'Sis would correlate their semantic choices on multiple levels, most frequently with the myth and ritual, in a context of stylistic play (1995: 285). Gonda notes in Vedic Literature: SaMhitaas and BraahmaNa's (1975: 65f.) that the R'Sis would take their revelation and transpose it into a verbal form marked and shaped by tradition and the canon as it was handed down and stored in memory from previous times. Elizarenkova expands upon this "handed down" criteria, characterizing it as a palette that includes syntax, metrics, sound, vocabulary, and morphological hues that can be woven and reworked according to the demands of expression for a given revelation. This database of stylistic elements was selectively drawn upon in the dynamic utterance of the mantras and later in the chosen tapestries of their repetitions in sacrificial or meditative contexts.

    In the research for this dissertation, I have constructed and employed a "cyber R'Si " of sorts. While the hybrid term may be novel, the reality is valid: the hard drive on the computer or WWW server has memorized the

shruti, and I draw upon this database to examine or explain a topic as would a R'Si in active thought (e.g., brahmodya or yajña recitation), through the HTML coding programmed into the e-text. In this case I have tried to create something akin to the Yaajñavalkya of BAAU 4 with regard to the self in Vedic and pre-UpaniSadic literature. However, as this study is an analysis of diachronic change in the terminological palette from which images of individuality were drawn in Vedic times, I have imposed a strict linear timeline of links upon this virtual Yaajñavalkya. As in the case of the RV, the order of the text as canonically transmitted and its actual historical sequence of composition differ greatly. There are a multitude of non-sequential segments--that is, for presentation in conventional analogue form. Thus, in reading and mapping with HTML the terminology related to the self in the RV I have linked each term according to the temporal--rather than textual--sequence of the passage. The RV is also cross-linked with the ShB.25 I also use the VaidikapadaanukramakoSaH for terms either not included in the primary database, or for those texts not currently in electronic form.

    It is ironic that the most fundamental, enabling element of this study--the HTML database (if you can imagine, in spite of the forward-thinking members of my committee and the Graduate College, my actual defense was a vestige of days arguably soon to pass: the technology did not even get mentioned, not once)--is found nowhere in the hardcopy pages for which it served as the resource. Here, in this electronic edition, however, the reader can access this text according to his/her interests--I have linked this dissertation intricately, freeing the reader from the constraints of my own chosen narrative sequence for recounting the investigation. Even more valuable is the inclusion of the entire databased, linked as needed throughout the study so that the reader can investigate passages and contexts for themselves, with the ease of a mouse click, without having to track down volumes, flip through pages, re-collate for historical sequence, placemark for cross-referencing, and so forth. All this is immediately available at any time throughout the study using either the JavaScript menu above in the control panel, or the links herein.


    The reader who has not used a resource like it cannot imagine what it is like to sit writing in the wee hours and find a word in a particular verse and be able to see--in less than 20 minutes (I have made an animation which, while large, shows how the RV-ShB web network lets me enter the world of the texts)--if that word occurs with the other hundreds of occasions where a related term occurs. The animation in the link here is continued below, viz. a discovery about how word use can reveal specific points of relative chronology as, for instance, between RV 10.27 and 10.34 (The Gambler Hymn). When I was working with bráhman, I wanted to know how many times words for speaking or utterance--verbs like -vac, -stu, and -gaa--were used with it. Bráhman occurs over 160 times in the early RV Family Books--MaNDala's 2-7--alone. Asking such a question with conventional research tools would require using the VaidikapadaanukramakoSaH (which may or may not be in the library--as at the University of Iowa--and costs over $200 for the early texts alone) for all occasions of bráhman, then all occasions of each root, cross-referencing each, then examining each passage to see if the verb form in any way modifies or acts upon bráhman. Once those occasions are listed, they must be put in historical order. With electronic resources, once I had mapped each occasion of bráhman, I can ask as many questions about the words used with it as I wish. Every time I do so, the results will always be pre-formatted in historical order--since this is how I mapped them--as well. Because I have linked the ShB to the RV, I can also check if any occasions where the RV mantra's are used in the ShB offer a different perspective. In

each case I am only investing about 1/2 hour to perform the inquiry. Computer technology is not some curious frill for adding flash and trendiness to scholarship. It opens the door to a multitude of questions that enable comprehensive study on a scale not possible before. The systematic and meticulous use of the conventional methodologies described above when ascertaining the answers maintains the quality of the academy amid the quantity of answered questions.