SEMANTIC CHANGE OF ĀTMÁN- IN THE R̥GVEDA AND THE ATHARVAVEDA

Abstracts

This paper studies the use of ātmán- in the R̥gveda and the Atharvaveda, and aims to show an ongoing semantic change through the period of composition and compilation of these texts. A cognitive and typological-functional approach is used here and all cases where ātmán- appears in the corpus have been analysed both grammatically and philologically. The results of this research show that ātmán- acquires new meanings and uses during the course of this process (‘being, ‘self’), without losing the previous ones (‘vital breath’, ‘breathe’). The systematisation of the different uses into a semantic map leads to the conclusion that: 1) the directionality of this semantic change is similar to that of equivalent words in other languages, and 2) this change is closely connected to that previously experienced by tanū́- ‘body, ‘self’. Thus, this research aims to provide specific evidence for cross-linguistic studies on grammaticalisation and the directionality of change.

Semantic change; Diachronic linguistics; Semantic maps; Vedic

Este artículo analiza el uso de ātmán- en el R̥gveda y el Atharvaveda y tiene por objetivo demostrar la existencia de una situación activa de cambio semántico a lo largo del período de composición de estos textos. El marco teórico utilizado es cognitivo y tipológico-funcional, y para llevar a cabo la investigación se ha analizado gramatical y filológicamente todos los pasajes en que se utiliza esta palabra dentro del corpus seleccionado. Entre los resultados más relevantes se encuentra que durante este proceso de cambio ātmán- adquiere nuevos valores (‘ser’, ‘sí mismo’), sin perder necesariamente los anteriores (‘principio vital’, ‘hálito’). La organización de los diferentes valores en un mapa semántico permite ver, en primer lugar, que la dirección del cambio semántico de ātmán- es similar a la experimentada por términos equivalentes en otras lenguas del mundo y, en segundo lugar, que este cambio se encuentra estrechamente relacionado por aquel experimentado previamente por tanū́- ‘cuerpo’, ‘sí mismo’. Así, el presente trabajo pretende ser un aporte individual a propuestas interlingüísticas acerca de la gramaticalización y la direccionalidad del cambio.

Cambio semántico; Lingüística diacrónica; Mapas semánticos; Védico

Introduction

Ātmán- is one of the most meaningful concepts found in the philosophical and speculative texts of ancient India, of particular relevance to Buddhism and Hinduism. The question of its essence, its intrinsic nature, has been the subject of innumerable research papers, as has been the difficulty in establishing a uniform meaning throughout the different periods of the Vedic language and classical Sanskrit (WILLMAN-GRABOWSKA, 1930; RENOU, 1997; GARDNER, 1998; ELIZARENKOVA, 2005). From a linguistic perspective, this difficulty is due to the changes experienced by the word through a continuum across the various periods of this Indo-Arian language as well as within the whole of the hymns that make up the corpus analysed in this research paper.

Particularly, an important antecedent for the present analysis has been the work of ELIZARENKOVA (2005), focussed on the values and usage of this word in the R̥gveda, that is, the most ancient collections of texts preserved in this language. The author claims that the meaning of this term in the earliest texts differs significantly from that found in the later texts. Trying to articulate the present work with that of ELIZARENKOVA (2005), the research objectives that originate this paper have been the following: a) generally, to obtain an overview of the different uses and meanings of this word in the R̥gveda and the Atharvaveda Śaunakīya (excluding data from the AV Paippalāda due to the more difficult access to complete editions as those existing for the Śaunakīya recension), hereinafter referred to as the RV and the AV respectively; b) and more specifically, to determine the way in which semantic changes occur from certain values to others in light of the existing methodologies in the study of the directionality of semantic change, grammaticalisation and semantic maps, in line with proposals by Hopper and Traugott (2003), Heine and Kuteva (2002), Haspelmath (2003) and François (2008).

This research will attempt to show, firstly, that certain semantic changes can be identified from the earliest texts of the RV to those more recent and on to the AV; and secondly, that these changes do not occur randomly but rather across contiguous zones of meaning. This paper will show that the semantic development of ātmán- is closely related to that of tanū́-, as the semantic changes of the latter appear to influence the changes in the former. Finally, given that one of the functions of ātmán- within the AV is to act as a reflexive marker, the acquisition of this role will be explained using semantic-syntactic criteria; this explanation is also directly linked to the development of this word in conjunction with that of tanū́-.

The research was conducted in the following phases:

1. All passages in the RV and the AV containing ātmán- have been compared and analysed to determine the most accurate meaning in each case.

2. The methodology of semantic mapping has been applied, analysing the process of semantic change according to a specific directionality in order to contrast the evidence with existing cross-linguistic proposals.

3. The change occurring in ātmán- has been compared to that of tanū́-, in order to identify similarities and differences.

In order to illustrate the main premise of this work, that semantic change occurs in a specific direction and across contiguous areas of meaning, this analysis is structured to present the different meanings independently, later relating them to each other using a semantic map. There are diachronic implications to the notion that semantic change necessarily occurs over time. However, it should be restated here that the principal aim of this work is not to argue in favour of a specific periodisation of the Vedic language, and therefore will not present the different uses of ātmán- in a chronological order.

Language of the RV and the AV

The RV and AV constitute the most ancient texts of any Indo-European language preserved in India, and as with other ancient languages of this family, it is a highly flexive language, with a very rich morphology and a large case system.

Generally speaking, the hymns of the RV and AV are liturgical texts, recited at different moments of the Vedic ritual. Many of these texts are repeated in both the RV and the AV (many hymns from one appear partially or entirely in the other), but it can be said that the RV contains a greater number of hymns alluding to rites of a public nature, while the principal texts of the AV are cosmogonic and theological hymns, used in saṃskāra rituals (“sacraments”), rites for women, rituals related to the king, prayers for health and long life, incantations for prosperity and luck, texts to atone for sins, texts to cure illness, and incantations against demons and sorcery (LUJÁN, 2010a); broadly speaking, it can be claimed that the AV largely consists of hymns of a private nature.

It is impossible to establish precise dates, but it is estimated that these hymns were composed between the middle of the second millennium B.C. and the end of the first millennium B.C. The most accepted chronology of these texts divides their composition into the following stages (WITZEL, 1989):

1. Early Vedic period, the central language of the R̥gveda-saṃhitā: RV 1.51-191, Books 2 to 7, that is, the “Family Books” (collections or maṇḍalas distributed according to priestly families to which each is attributed), 8.1-66 and perhaps parts of Book 9.

2. Later Early Vedic period, consisting of the language of RV Books 1.1-50, 8.67-103, Book 10 and perhaps parts of Book 9.

3. Ancient Vedic period, corresponding to the language of the late Mantra period, representing the language of the RV-Khila (a series of apocryphal hymns), Sāmaveda, Atharvaveda (Paippalāda and Śaunakīya), Yajurveda.

Some observations should be made regarding the periodisation of the AV and its relation to the RV. Firstly, there are not merely chronological differences between the hymns of the RV and the AV, but dialectal differences as well (PARPOLA, 2002; WITZEL, 1997). Secondly, the AV can be considered as later than the RV, not because of its antiquity but essentially because of the process of compilation and canonisation, closely linked to the most recent linguistic characteristics of some of its hymns (WITZEL, 1997).

Distribution of ātman- in the RV and AV

The term ātman- is of Indo-European origin and etymologically related to the German Atem ‘breath’ and the Gothic uz-anan ‘exhale’ (MAYRHOFER, 1986-1992, p. 164; PUDDU, 2005, p. 91). There is a wide disparity in the frequency of its use, appearing just 22 times in the RV, and in 48 passages in the AV. The distribution of its use according to the language periods indicated above is as follows:

Table 1
– Distribution of ātman- in the RV and AV

	RV 2-7	RV 8-9	RV 1 & 10	AV
Nº of passages	2	5	15	48

• Source: Author’s elaboration, based on Lubotsky (1997) and Whitney (1881).

The conclusions drawn here cannot be considered absolute but merely an approximation as there is no precise word count for each specific case. But, given the current lack of any quantitative database of the frequency of word use in the RV and the AV, it is not possible to provide more information at present. However, with this caveat, it may be instructive to note the distribution of its endings:

Table 2
– Distribution of ātman- in the RV and AV by morphology

	N.SG	AC.SG	LOC.SG	ABL.SG	INS.SG	DAT.SG	LOC.PL	TOTAL
RV	15	4	1	2	-	-	-	22
AV	12	12	8	4	6	3	3	48
TOTAL	27	16	9	6	6	3	3	70

• Source: Author’s elaboration.

Although merely exploratory, these figures point to a possible extension from direct towards oblique cases in the AV and not in the RV, which has both morphological and syntactic implications.

Uses and meanings

Traditionally, scholars have highlighted the polysemy of this term. Among the most frequently attributed meanings are ‘vital breath’, ‘spirit’, ‘being’ or ‘soul’, although this concept should not simply be understood as separate from the physical and corporal aspect of one’s being. Thus, notes Grassmann (1976, author’s italic):

1)Hauch tritt mit der ausdrücklichen Parallele vāta klar hervor (…); mit ihr in naher Berührung steht, 2) Athem, Odem, Lebenshauch; weiter 3) Lebensgeist, Lebensprincip, auch 4) vom Geiste der Krankheit (yákṣmasya) wird es einmal gebraucht; 5) der lebendige Leib, als Einheit aufgefasst.¹

In addition to the different lexical values mentioned above, related literature often makes reference to its possible reflexive use (MENDOZA, 1984; KULIKOV, 2007; HETTRICH, 2010) although there is no consensus among specialists in this matter. Mayrhofer (1986-1992), for example, believes this value already occurs in the RV while other authors, such as Delbrück (1888), maintains that it only appears from the AV on.

Additionally, and in relation to the question of its essence, in studying the concepts referring to the English term self, Gardner (1998) maintains that ātmán- is related to a vital and dynamic element that remains to a degree unspecified. Thus, the term is closely connected to elements characterised by their animacy features.

Considering the semantic description found in standard reference texts, and based on a painstaking analysis of all passages in which the term occurs, the following panorama can be offered, classifying the different meanings and uses encountered, first in the RV and later in the AV.

R̥gveda

Vital principle

This is the most frequently cited meaning and is found throughout the entire corpus. Translations may vary between the notions of ‘spirit’, ‘vital breath’ or ‘vital principle’, but what appears clear is that this concept makes reference to some element related to respiration and from this to life itself (ZYSK, 1993). Using this meaning, the texts make reference to that which is full of life or vitality in the most recent Books (examples 1 & 5), as well as the two sole cases found in the “Family Books” (examples 2 & 3) and in more complex Books such as 9 (example 4):

(3) should not be interpreted to mean that the vital principle belongs to something that is not alive but rather that inanimate objects also have their own vital principle. A similar case appears in (2), as the wind, an element of nature, is deified as an animated element. From this, it seems logical to apply this concept to elements which are initially not alive but animated. For example:

Within the scope of animated elements, this term serves for both human and divine beings:

Given the paucity of evidence in the most ancient texts of the RV, it is impossible to affirm categorically that the term was initially used to refer to animated beings and was then extended to inanimate objects. However, this hypothesis cannot be ruled out as this direction is closely related to the proposed hierarchy of animation in many languages (CROFT, 2003).

While this meaning is present throughout the entire RV, it is only in the most recent hymns that this vital principle is identified as one of the parts of the being and, as such, in a close relationship with other parts of the being. But this particularity may merely be due to a lack of evidence, for example:⁵

Ásu- and ātmán- are two components of the being that have a strong interrelation (KAHLE, 2012) and both are involved in this process of semantic change. Some translations present the two as almost synonyms although there is a fundamental difference found in these texts: ātmán- is located within the subject, while ásu- is a state into which the subject enters; thus, ásu- is an element associated with death and the Afterlife, at times approaching the notion of ψυχή (ELIZARENKOVA, 2005, p.127); in other words, the concept of ásu- corresponds to that part of life that extends beyond death. It is no coincidence that these types of references appear in Books 1 and 10, which are precisely those which begin alluding to death and scatology.

Another element of the being linked in these texts with ātmán- is tanū́- ‘body’:⁶

The parallel between these clauses suggests a difference between ātmán- and tanū́-: the character of “holder” or “container” of tanū́- (“dwelling”, a place where one resides) is contrasted with the character of “content” or “contained” of ātmán- (“food” or “nourishment”, which enters into the dwelling and fills it with life).

As noted by Elizarenkova (2005), the relation between this concept and words related to ‘food’ or ‘nourishment’ in the RV should come as no surprise; it also occurs in RV 7.101. 6 (pitú-) and in RV 9.85.8 (dhāsí-). This frequent relation leads the author to suggest that the vital principle may be understood as something physical in relation to the human body (ELIZARENKOVA, 2005). However, we must recall, firstly, the presence of this concept in relation to animated elements (although not necessarily animate), and, secondly, the importance of ingestion in Vedic ritual, not simply as a habitual action but as the performance of a religious rite (MALAMOUD, 1996): the priest can be regarded as a cook and the oblations are the substance that is prepared and consumed. The main gods of the Vedic pantheon are invited to eat and drink, and are petitioned to lend their strength and actions to those making the offering (PATTON, 2005). Ingestion is essential to the rite and for relations of reciprocity and the food can also be considered as the “essence” of the sacrifice, which explains the parallels between these two kinds of concepts.

From the idea that ātmán- refers to what is full of life, it seems that in certain contexts the meaning has shifted from “that which fills with life” to “the being with life”, including both corporeal and incorporeal aspects. This becomes especially evident in passages that mention the extraction of an illness from ātmán-. In other words, a change takes place due to metonymy, in which the point of departure is that ātmán- is that which fills with life and the finishing point is that ātmán- is a being full of life, an animated being.

Animated being

The translation of the concept of ‘being’ is highly complex and has given rise to a great deal of discussion and debate in both Western and Indian thinking. The concept of ātmán- lies at the heart of this debate, although it is difficult to maintain that in the RV and AV there is the same degree of ontological speculation. On the contrary, within this corpus ‘being’ can be understood fundamentally as the idea of the unity of the being, connecting the corporeal and incorporeal aspects, either of oneself or of another. Most probably, given the emphasis found in the following passages, a better translation could be ‘sameness’. It is in this sense that the following passages should be understood although the translation as ‘being’ has been kept for the sake of clarity:

In (8) and (9), the value of ātmán- goes beyond the idea of ‘vital breath’ and involves the corporeal aspect of the being. This is not the case in those few instances that can be included within the Family Books of the RV, nor in a large number of the remaining passages. Note that (8) forms part of the texts containing medicinal incantations, a frequent occurrence especially in the AV. Example (9), however, is more difficult to place since Book 9 is a later compilation of hymns dedicated to Soma. For reasons to be discussed below (see “Tanū́- and ātmán-”), these can be considered as later texts.

To provide a broad summary of the uses of ātmán- in the RV, one could say it is principally used to mean ‘that which fills with life’, ‘vital breath’ or ‘vital principle’, referring to human beings, to gods or to certain animated elements such as the wind. This use can be found in 16 of the 22 passages: 7.87.2a, 7.101.6b, 8.3.24ab, 9.2.10d, 9.6.8a, 9.85.3b, 10.92.13c, 10.97.11c, 10.107.7c, 10.168.4a, 1.34.7d, 1.73.2d, 1.115.1d, 1.162.20a, 1.163.6a and 1.164.4c. In the majority of passages, this term is used in direct cases (nominative and accusative).

In some of these cases, especially when it is connected to other key concepts such as tanū́- or ásu-, the term ātmán- makes reference to one of the parts of the being. In contrast, it refers to the self or being in a more general sense in those passages in which it makes reference to the extraction of an illness or part of a funeral hymn (situations which are less typical in the oldest hymns): 10.16.3a, 10.97.4d = 8d, 10.163.5c =6c and also 9.113.1c.

Atharvaveda

Vital principle

Both of the values appearing in the RV are present in the AV, particularly that of ‘vital breath’, ‘vital principle’:

As occurs with the most recent hymns of the RV, and more frequently in the AV, ātmán- is combined with other elements referring to the animated being, as for example with ásu-, and also with tanū́-:

Animated being

As in the RV, there are passages in the AV in which the concept does not refer to just one of the components of the animated being, but to the being as a whole, in which the corporeal or material aspects converge with the incorporeal aspect. This is in line with the observations of Gardner (1998) regarding the degree of imprecision or non-specificity of ātmán- in certain contexts. Among the cases in which ātmán- is combined with other elements referring to the whole of the animated being, are the following:

The genitive of (13) reveals a distinction between tanū́- and ātmán-, as “container” and “content”. Additionally combinations with another key element prāṇá- are found in the AV:

This term, prāṇá-, merits particular mention given its semantic proximity to ātmán-, which can perhaps be explained by a common etymology (an- ‘breathe’ + prefix prá) (MAYRHOFER, 1986-1992). The appearance of prāṇá- is notably infrequent in the most ancient texts of the RV, only occurring on five occasions (KAHLE, 2012), four of which belonging to the most recent texts (ELIZARENKOVA, 2005). It can also be noted here that, as opposed to those passages in which ātmán- appears (and thus, the “content”), here this meaning is expressed by prāṇá-, displacing the other concept of “container”.9

Considering the etymological relation and the internal chronology of the texts, this may explain the lack of interrelation between the two concepts in the earliest texts. As noted by Elizarenkova (2005), if the common etymology had existed at the time of composition of the earliest texts, it may have been difficult to consider both elements as differentiated components of the being. In the AV, however, this distinction can be made through the loss of this semantic association (ELIZARENKOVA, 2005). If we regard this as a possible semantic development, it seems logical that this distinction would be even greater in the brāhmaṇas, generally later than the RV and a large part of the hymns of the AV. This would be the case to the degree to which prāṇá-¹⁰ assumes the ancient meaning of ātmán- as the ‘vital breath’ when ātmán- comes to be specifically associated with the concept of ‘animated being’, ‘selfhood’, “assuming the aspect of ásu- as the spiritual component of man that transcends death” (KAHLE, 2012). In fact, this appears to be the meaning pointed to in example (14).

Interior being

Taking the idea of ‘full of life’ as a point of departure, it is possible to reach the notion of ‘being (full of life)’, as seen in the RV. In the AV, this semantic change appears to go further towards encompassing an idea that distinguishes between the two aspects of the being: the interior and exterior. This is absolutely crucial if we recall the conception of ātmán- as that which is able to enter into the being:¹¹

As can be seen, this is also closely related to the idea of the interior and the animated being understood as a whole, that is, including the corporeal and incorporeal aspects:

Ātmán-, understood as the ‘interior being’ is now in opposition to the ‘exterior being’:

Rūpā́ni and ātmā́nam are antithetical: rūpā́ni ‘the outer form of things’; ātmā́nam ‘thy own nature’. It is a controlling characteristic of Vedic conceptions that the inner, true nature of any divinity, or instrument of power, must be understood in order to control its influence or power. (BLOOMFIELD, 1897, p. 402-403).¹³

This may gradually lead to a reflexive interpretation, as is evident in the following passage:

At this point is seems clear that the interior and that which forms an intrinsic part of the being, its quintessence, are merged:

Reflexive marker

Through the notion of ‘being’ understood as a whole (corporeal and incorporeal aspects), it is a simple matter to extend its use to that of a reflexive marker, as seen when ātmán- is co-referent to another element in the clause:

According to the data obtained here, it is difficult to support that the reflexive meaning is already present in the RV given that there is only one passage that exhibits the properties of a reflexive (RV 9.113.1c), attributed to the most recent period.

Generally speaking, cross-linguistic studies of reflexivity largely focus on direct cases, that is, those in which co-reference is established between the first and the second actants, typically subject and direct object (accusative case). However, the phenomenon of reflexivity can be extended to indirect cases as well, that is, when there is a third actant, generally in oblique cases, as in (19) and (20).

Examples from different languages help to understand how this term comes to be interpreted as a reflexive, since there are plenty of cross-linguistic evidence for reflexives that develop from elements meaning ‘spirit’, ‘being’, ‘person’, ‘body’ (SCHLADT, 2000; KÖNIG; SIEMUND, 2000; HEINE; KUTEVA, 2002), with possible intermediate stages in which the element intensifies the co-reference. In fact, this emphatic value can also be found in passages of the AV, emphasising co-reference (fundamentally with accusative) and in situations of nominal intensification (generally oblique cases, especially instrumental and locative), in which ātmán- produces an effect similar to that found in the emphatic expressions “himself/herself”, “in person” or “in the flesh”:

An interpretation merely based on the lexical sense of the word would be very peculiar: ?‘Becomes by his own vital breath’, ?‘May he reach the curve with his vital breath’. But if we consider an emphatic reading applied to the idea of the whole being, similar to what occurs with the emphatic expression “him/herself”, the text becomes much clearer. Further, it is consistent with the development of a reflexive grammatical element, which often develops through the need to intensify the co-reference between two elements in a clause.

In summarising the uses in the AV, we find, firstly, that many are maintained in the RV in certain contexts; secondly, that with this meaning the combination of ātmán- with other concepts referring to parts of the being are more frequent; and thirdly, there is a degree of expansion in the range of possible meanings (‘being’, ‘interior being’, reflexive and emphatic).

Although many passages are more difficult to classify in a single meaning (there are many ambiguous cases), the following distribution of passages according to meaning can be offered:

‘Vital principle’: 13.2.35d, 5.1.7b, 11.8.31cd, 5.5.7cd, 5.9.7ab, 12.2.34cd, 3.29.8cd, 10.8.44cd, 5.9.7cd, 19.51.1ab, 9.9.4cd, 16.3.5, 9.4.10ab, 7.111.1, 7.67.1ab, 6.53.2ab.

‘Being’: 1.18.3a, 7.57.1cd, 11.5.22ab, 19.48.5cd, 9.1.11-13cd, 9.1.16cd, 8.6.13ab, 3.15.7cd, 11.5.15cd, 5.29.5ab, 4.18.6cd, 19.33.5ab, 18.2.7ab, 6.16.2cd, 4.9.7cd.

‘Interior being’: 9.8.9cd, 7.53.3cd, 4.12.2b, 4.20.5cd, 12.4.30ab.

Reflexive: 15.10.2a, 12.1.10cd, 15.1.2, 12.3.54ab, 9.6.21, 19.17.1-10cd, 12.3.51cd, 9.5.30ab, 16.7.5ab.

Emphatic: 9.5.31-36ef, 8.2.8cd, 5.29.6-9c.

A semantic map

The use of ātmán- throughout the corpus points towards an increasing frequency of use and an extension of the range of contexts from the earliest hymns of the RV to the series of hymns in the RV 10 and on to the AV. Although it is beyond the scope of the present work, it would be interesting to study the continuation of this extension into later texts, possibly into Vedic prose, in order to confirm the affirmations of several authors as to the consolidation of a reflexive meaning (GARDNER, 1998; KAHLE, 2012).

This process of change connects their synchronic and diachronic aspects. On one hand, there is a clear set of differences between the three nuclei that are not only dialectal but also of content. The hymns of the RV and AV where we find an extension in the range of uses are often related to the practice of medicinal and funeral incantations. Furthermore, although not a definitive or irrefutable argument, the differences found may point towards a relative chronology where the semantic changes occurring over time in the three nuclei generally coinciding with the broad phases of these texts. The first phase generally coincides with the most typical meaning of ‘vital breath’ or ‘vital principle’; the second phase, with texts where the meaning extends to ‘being’, ‘person’, in general terms, especially in Book 10 of the RV, and the most recent, the third phase, corresponds to the AV, where there is an extension to a much broader range of uses (‘interior being’, emphatic, reflexive) and towards the rest of the nominal paradigm. This semantic change may be related to a morpho-syntactic development as in the majority of cases the meaning ‘vital principle’ are direct cases (nominate and accusative), while the other meanings are generally associated with indirect cases.

It is essential to identify different meanings and their interrelation in order to understand the process of semantic change. This process is not random but occurs across contiguous zones of meaning. A useful tool for analysis is a graphic representation using a semantic map that also allows synchronic aspects to be linked to diachronic aspects. A semantic map can be defined as “A geometrical representation of functions in a ‘conceptual/semantic space’ interrelated through connecting lines, thus creating a network.” (HASPELMATH, 2003, p.213). That is, a semantic map represents a selection of meanings and their interrelation and thus helps to create a network of shared semantic features. In order to understand the interrelation between different meanings, it is essential to understand the properties of each. As an example, Haspelmath (2003) proposes the following semantic map applied to the development of the reflexive:

Figure 1
– Semantic map of the reflexive and middle functions.

Although one may object to certain aspects of this map, particularly the development of grooming actions from full reflexives and not the reverse (see SCHLADT, 2000), it serves to show that semantic change does not occur randomly but rather through the shared features of contiguous functions. Thus, the contiguity between the emphatic reflexive and full reflexive is connected to a simple change from one meaning to another and not a large jump from the emphatic reflexive to an anticausative.¹⁴

The following semantic map¹⁵ has been created based on the evidence of the uses and functions of ātmán- in the RV and AV:

Figure 2
– Semantic map of ātmán- in the RV and AV

This map reflects the semantic change produced from one set of hymns to another. In the process, the term does not acquire new meanings in detriment to others but simply adds an additional layer of meaning. This change occurs across contiguous zones of meaning. In this case, the direction of the change is from ‘vital principle’ to ‘animated being’ and from there to other values due to the shared features of the contiguous zones of meaning. It would be difficult to explain an emphatic use, for example ‘the person themselves’ developing from the idea of ‘vital breath’, without an intermediate point of meaning as ‘the person as a whole’. Similarly, it is unexpected that a reflexive value would develop from the idea of ‘vital principle’ and towards a later meaning of ‘animated being’.

The direction of change seen here fully coincides with proposals for a similar change applied to various other languages with a similar development of reflexives from vocabulary related to the soul, the body or its parts, and through intermediate emphatic meanings which arise through a need to create co-reference (SCHLADT, 2000; KÖNIG; SIEMUND, 2000; HEINE; KUTEVA, 2002). This change is generally categorised as a case of grammaticalisation, that is, the term passes from a purely lexical element to having, in certain contexts, a grammatical function. During this process of grammaticalisation, contexts frequently become more generalised, that is, the element expands its range of meaning and possible contexts for use, resulting in cases of polysemy (HOPPER; TRAUGOTT, 2003), as shown in this research.

With our semantic map of ātmán- in the RV and AV, it can be instructive to compare this to the map of ātmán- in classical Sanskrit by FRANÇOIS (2008), in which the data are not drawn directly from the texts but from secondary literature:

Figure 3
– Semantic map of ātmán- in Sanskrit.

This map also indicates a direction of change across contiguous features of meaning. The reflexive does not develop directly from the idea of ‘vital breath’, but through a series of intermediate points, the most essential of which is ‘the person themselves, the own being’. The development of the meaning ‘soul’ from the idea of ‘life force’ is also coherent within this process, although this value was not found in the texts of the RV and AV.

Note, however, that there are some significant differences between the two maps. These differences demonstrate the importance of philological research for a rigorous analysis of a language that, in turn, allows for more effective cross-linguistic comparisons. Firstly, when dealing with an ancient language as in this case, research that does not include a detailed study of the primary source material is less precise: the Sanskrit found in reference works tends to have an excessively broad temporal margin that may obscure the data that can be gathered from the diachronic process. Note that several of the meanings in this map are not really found in the texts of the RV and AV while the distinction between ‘breath of life’ and ‘life force’ can seem rather unnecessary.

Secondly, a detailed philological work can more exactly determine intermediate values to be taken into account in semantic development. Particularly, the RV and AV reveal the essential intermediate reading of ‘person’, understood both in its corporeal and incorporeal aspects, allowing the development of the meaning ‘the own self’, ‘oneself’, ‘him/herself’.

Ātmán- and tanū́-

Beyond the cross-linguistic evidence, we can look to possible internal factors affecting the semantic change in ātmán-. As shown, this word is generally associated with others that also make reference to parts of the being and which may at times share a similar meaning. This is particularly the case with tanū́-, a feminine noun originally meaning ‘body’ that has also undergone semantic and syntactic changes in the corpus, moving from a much more lexical function to that of a reflexive marker. While it is impossible to provide a detailed analysis of this word and its semantic changes here, this subject has been exhaustively explored in a number of works (PINAULT, 2001; HOCK, 2006; KULIKOV, 2007; HETTRICH, 2010; ORQUEDA, 2013).

Succinctly put, the lexical meaning of tanū́- as ‘body’ is predominant, especially in the earliest texts and continues to be used throughout the corpus:

From the most recent books of the RV on, we find the addition of the meaning of ‘being’, also understood in an integral manner, as a conjunction of the corporeal and incorporeal aspects. For this reason, Gardner (1998) chooses to translate this word as ‘presence’.

The term later acquires a reflexive value (25, 26), but before reaching this point it is found in a number of different contexts in which its interpretation lies ambiguously in between a lexical and a reflexive value (24):

During the course of development towards a reflexive use, there are frequent contexts in which tanū́- functions as an intensifier. This is due to the familiar pattern of semantic change in which an intermediate stage of intensification is required:

In (26) and (27), tanū́- is used as a reflexive and, in both cases, this interpretation arises from an intermediate emphatic situation: a simple reflexive interpretation can be achieved from the use of the middle voice. Thus, tanū́- (especially when reinforced with svá-) intensifies the situation of co-reference.

The existence of ambiguous (for example, lexical and reflexive) passages is fundamental to proposing the possibility of semantic change (SCHLADT, 2000). In fact, in very few cases (and never in the earliest hymns of the RV) can tanū́- be understood solely as a reflexive, with a complete loss of its lexical meaning. However, ambiguous passages in which both interpretations are possible do appear in the RV, raising the possibility that the process of semantic change undergone by tanū́- towards becoming a reflexive marker with an intermediate stage with an emphatic nuance is older than the similar process undergone by ātmán-. The process of semantic change undergone by tanū́- may have led to a superimposition over ātmán- for certain meanings and a close relation between both terms. Thus, semantic change may have occurred to ātmán, partly as a result of the model provided by tanū́-. Furthermore, this may explain how some of the uses of tanū́- are assumed by ātmán- while some uses of ātmán- are assumed by prāṇá-.

Returning to (9), the only passage in the RV in which ātmán- approaches a reflexive use, this may be explained as a late case and repeat of what occurs with tanū́- as similar earlier contexts for this word can be found. Note the parallel with the following passages from Book 3:

In all of these cases Agni or Indra is invoked and these gods are related to the capacity to give (or give oneself) strength. While this semantic and syntactic structure appears with tanū́- in the earliest texts of the RV, a similar situation with ātmán- is found in later texts and with a corporeal meaning added, precisely the lexical value originally attached to tanū́-.

Consequently, one could offer the hypothesis that the passage with ātmán- was in some way constructed using structures proper to tanū́-, after assuming part of its meaning. This is not surprising if one considers the process mentioned previously whereby the notion ‘that which fills with life’ evolves into ‘the being full of life’, and then finally to ‘being’, given that tanū́- also changes from the idea ‘body of a being’ towards simply ‘being’ through metonymy, with the same observations made for ātmán- regarding the ontological problem of translating this term. An additional argument for this hypothesis may be the fact that in (9) ātmán- is in the locative, as occurs in (28) and (29). Interestingly, the use of oblique cases of this word can be found relatively late. Furthermore, there is a clear ambiguity in (9) between a lexical interpretation and a reflexive, which relates it perhaps even more closely to the development of tanū́- towards a reflexive value through ambiguous situations.

Final comments

The results of this research show that, in line with current cross-linguistic proposals, ātmán- is not limited to a single meaning but includes a range of different meanings, such as ‘being’, ‘vital breath’, ‘oneself’, etc. But this range of meanings cannot be explained solely as a case of polysemy at a synchronic level, as this semantic change appears to be related to diachronic development throughout the hymns of the RV and AV. Although there is limited data for the earliest RV language (Books 2 to 7), one can postulate a certain diachronic variation connected to the emergence of specific features and semantic meanings in each stage:

1. Early RV, the oldest nucleus of the RV and part of the most recent, in which the principal meaning is that of ‘vital breath or vital principle’, ‘breath’.

2. Late RV, consisting of some more recent hymns from the RV and part of the AV, in which the term assumes a meaning similar to ‘interior being’ and ‘animated being’ or ‘person’. Through this, we can interpret a degree of emphasis on the idea of the self, similar to the meanings attributed to tanū́-, which appears to indicate the uses of tanū́- being partially assumed by ātmán-.

3. Finally, and in parentheses given the difficulties in establishing the chronology of the AV, the greatest part of this collection can be classified as a third stage, understood as a series of passages in which the word acquires a more clearly reflexive value. This is possible through the assumption of the meanings recently attributed to tanū́- and may explain the fact that once ātmán- fully assumes a reflexive role, to be continued into future texts, tanū́- stops its incipient development towards the role of a reflexive marker and never become fully grammaticalised.

The texts provide vivid evidence of semantic change. In order to be fully understood, it requires an analysis of the interrelation between each of the elements during the different periods of the language. This also leads to a reflection on diachronic and synchronic studies: a change in one of these aspects is intrinsically linked to changes in the other. In turn, a diachronic change in one of the elements of a language may result in a lexical reorganisation at the synchronic level.¹⁹

Finally, and in defence of philology, it is evident that an exhaustive analysis of the text is essential to providing a precise linguistic description and must be taken into account in the field of linguistic typology.

Abbreviations

ABL Ablative IMPV Imperative PF Perfect ACC Accusative INJ Injunctive IP Interrogative Particle ACT Active Voice INS Instrumental PL Plural AOR Aorist INT Intensive PRS Present DAT Dative LOC Locative PTC Participle DEM Demonstrative MED Middle Voice REL Relative G Genitive OPT Optative SG Singular

ABL	Ablative	IMPV	Imperative	PF	Perfect
ACC	Accusative	INJ	Injunctive	IP	Interrogative Particle
ACT	Active Voice	INS	Instrumental	PL	Plural
AOR	Aorist	INT	Intensive	PRS	Present
DAT	Dative	LOC	Locative	PTC	Participle
DEM	Demonstrative	MED	Middle Voice	REL	Relative
G	Genitive	OPT	Optative	SG	Singular

BIBLIOGRAPHY

AUFRECHT, T. Die hymnen des R̥igveda. Bonn: Adolph Marcus, 1877.

GELDNER, K. Der Rig-Veda aus dem sanskrit ins deutsche übersetzt und mit einem laufenden kommentar versehen. Cambridge: Harvard University Press, 2003.

MAX PLANCK INSTITUTE FOR EVOLUTIONARY ANTHROPOLOGY. Departamento de Lingüística. Leipzig Glossing Rules. Leipzig, 2008. Available in: <http://www.eva.mpg.de/lingua/resources/glossing-rules.php>. Access in: 29 jul. 2014.

VANHOVE, M. (Ed.). From polysemy to semantic change: towards a typology of lexical semantic associations. Amsterdam: John Benjamins, 2008.

WHITNEY, W. D. Atharva-veda saṁhitā. Cambridge: Harvard University Press, 1905.

WITZEL, M. Substrate languages in old Indo-Aryan (Ṛgvedic, Middle and Late Vedic). Electronic Journal of Vedic Studies, v.5, n.1, p.1-67, 1999. Available in: <http://www.ejvs.laurasianacademy.com/ejvs0501/0501ART.PDF>. Access in: 11 ago. 2012.

Acknowledgements

This research was conducted within the framework of Research Project FFI2012-36069-C03-02: “Estudios de formación nominal: lenguas paleohispánicas e indoeuropeas antiguas” of the Department of Greek Philology and Indo-European Linguistics of the Complutense University, Madrid.

I would like to express my sincere gratitude for the close reading and comments of Óscar Figueroa Castro, and the invaluable contributions of Eugenio Luján Martínez and Julia Mendoza. This work was particularly enriched thanks to the pertinent observations of the peer reviewers to whom I am very grateful. All errors are my own.

REFERENCIAS

• BAUM, D. The valency and meaning of the verb yaj in the R̥g-veda. In: SHULMAN, D. (Ed.). Language, ritual and poetics in ancient India and Iran: studies in honor of Shaul Migron. Jerusalén: The Israel Academy of Sciences and Humanities, 2010. p.3-15.
• BLOOMFIELD, M. Hymns of the Atharva-Veda.Oxford: Oxford University Press, 1897. (The Sacred Books of the East, v.42).
• DELBRÜCK, B. Altindische syntax Halle: Verlag der Buchhandlung des Waisenhauses, 1888.
• CROFT, W. Typology and universals Cambridge: Cambridge University Press, 2003.
• ELIZARENKOVA, T. The word ātmán in the R̥gveda: Indologica Taurinensia, Turín, v.31, p.121-134, 2005. Disponible en: <http://www.indologica.com/index-volume31.asp> Acceso en: 29 jul. 2014.
  » http://www.indologica.com/index-volume31.asp>
• FRANÇOIS, A. Semantic maps and the typology of colexification: interwining polysemous networks across languages. In: VANHOVE, M. (Ed.). From polysemy to semantic change: towards a typology of lexical semantic associations. Amsterdam: John Benjamins, 2008. p.163-215.
• GARDNER, J. R. The Vedavid database and dissertation on the developing terminology for the self in Vedic India. 1998. Disponible en: <http://vedavid.org/diss/index.html> Acceso en: 12 abr. 2014.
  » http://vedavid.org/diss/index.html>
• GRASSMANN, H. Wörterbuch zum Rig-Veda Wiesbaden: Otto Harrassowitz, 1976.
• HASPELMATH, M. The geometry of grammatical meaning: semantic maps and cross-linguistic comparison. In: TOMASELLO, M. (Ed.). The new psychology of language Nueva York: Lawrence Erlbaum Associates Publishers, 2003. v.2, p.211-243.
• HEINE, B.; KUTEVA, T. World lexicon of grammaticalization Cambridge: Cambridge University Press, 2002.
• HETTRICH, H. Tanú- als reflexivpronomen im R̥gveda? In: FINCKE, J. (Ed.). Festschrift für Gernot Wilhelm anläßlich seines 65. geburtstages Dresden: Islet, 2010. p.175-183.
• HOCK, H. H. Reflexivization in the Rig-Veda (and beyond). In: TIKKANEN, B.; HETTRICH, H. (Ed.). Themes and tasks in old and middle indo-aryan linguistics. Delhi: Motilal Banarsidass, 2006. p.19–44. (Papers of the 12th World Sanskrit Conference, v.5).
• HOPPER, P.; TRAUGOTT, E. Grammaticalization Cambridge: Cambridge University Press, 2003.
• KAHLE, M. D. El surgimiento de la doctrina de la transmigración de las almas en la India. Madrid: Servicio de Publicaciones de la Universidad Complutense de Madrid, 2012.
• KÖNIG, E.; SIEMUND, P. Intensifiers and reflexives: a typological perspective. In: FRAJZYNGIER, Z. (Ed.). Reflexives: forms and functions. Ámsterdam: John Benjamins, 2000. p.41-70.
• KULIKOV, L. The reflexive pronouns in Vedic: a diachronic and typological perspective. Lingua, Amsterdam, n.117, p.1412-1433, 2007.
• LUBOTSKY, A. A R̥gvedic word concordance New Haven; Connecticut: American Oriental Society, 1997.
• LUJÁN, E. R. Textos mágicos del Atharvaveda In: ÁLVAREZ-PEDROSA, J. A.; TORALLAS TOVAR, S. (Ed.). Edición de textos mágicos de la Antigüedad y de la Edad Media Madrid: CSIC, 2010a. p.43-59.
• LUJÁN, E. R. Semantic change. In: BUBENIK, V.; LURAGHI, S. (Ed.). Continuum companion to historical linguistics. Londres: Continuum Publishing Corporation, 2010b. p.286-310.
• MALAMOUD, C. Cooking the world Delhi: Oxford University Press, 1996.
• MAYRHOFER, M. Etymologisches wörterbuch des Altindoarischen Heidelberg: C. Winter. v.1-v.2, 1986-1992.
• MENDOZA, J. La reflexividad en indoeuropeo. In: BERNABÉ, A. et al. (Ed.). Athlon: satura grammatica in honorem Francisci R. Adrados. Madrid: Gredos, 1984. p.325-338.
• ORQUEDA, V. La expresión de la reflexividad en la lengua védica 2013. 513f. Tesis (Doctorado en Estudios del Mundo Antiguo) - Facultad de Filología, Universidad Complutense de Madrid, Madrid, 2013.
• PARPOLA, A. From the dialects of old Indo-Aryan to Proto-Indo-Aryan and Proto-Iranian. In: SIMS-WILLIAMS, N. (Ed.). Indo-Iranian languages and peoples. Londres: The British Academy, 2002. p.43-102. (Proceedings of the British Academy, v.116).
• PATTON, L. Bringing the gods to mind: mantra and ritual in early Indian sacrifice. Berkeley: University of California Press, 2005.
• PINAULT, G. Védique tanú- et la notion de personne en indo-iranien. Bulletin de la Société de Paris, Paris,v.96, n.1, p.181-206, 2001.
• PUDDU, N.Riflessivi e intensificatori: greco, latino e le altre lingue indoeuropee. Pisa: ETS, 2005.
• RENOU, L. The word ātmán In: BALBIR, N.; PINAULT, G.-J. (Ed.). Choix d’études indiennes Paris: École francaise d’Extrême-Orient, 1997. p.877-883.
• SCHLADT, M. The typology and grammaticalization of reflexives. In: FRAJZYNGIER, Z.; CURL, T. S. (Ed.). Reflexives: forms and functions. Ámsterdam: John Benjamins, 2000. p.103-124.
• WHITNEY, W. D. Index verborum to to the published text of the Atharva-veda New Haven: American Oriental Society, 1881.
• WILLMAN-GRABOWSKA, H. L’idée de l’Ātmán du Rig-Veda aux Brāhmana. Rocznik Orientalistyczny, Varsovia, v.2, p.10-25, 1930.
• WITZEL, M. The development of the Vedic canon and its schools: the social al political milieu. In: WITZEL, M. (Ed.). Inside the texts, beyond the texts: new approaches to the studies of the Vedas. Cambridge: Cambridge Univesrity Press, 1997. v.2, p.257-334.
• WITZEL, M. Tracing the Vedic dialects. In: CAILLAT, C. (Ed.). Dialectes dans les litteratures indo-aryennes París: Collège de France, 1989. p.97-264.
• ZYSK, K. The science of respiration and the doctrine of the Bodily Winds in ancient India. Journal of the American Oriental Society, Ann Arbor, v.113, n.2, p.198-213, 1993.

• 1
  “Breath clearly occurs with the explicit parallel vāta (...); and is in close contact with this, 2) vital breath, respiration, breath of life; even 3) spirit of life, vital principle, and 4) it is used once as spirit of the illness (yákṣmasya); 5) the living body, considered as a whole.” (GRASSMANN, 1976, our translation)
• 2
  Throughout the work only ātman- is not translated in the glosses to avoid conditioning the semantic interpretation in each case.
• 3
  Dákṣiṇā is the remuneration or payment to the priest paid by those requesting the rite or sacrifice to be performed.
• 4
  The use of the masculine here despite dákṣiṇā being feminine may be a case of attraction by ātmán-
• 5
  Similar to AV 9.9.4c.
• 6
  Given that this article is focussed on the semantic behaviour of ātmán- rather than tanū́-, the latter will be glossed in each case.
• 7
  Name of an illness, possibly a type of tuberculosis (ELIZARENKOVA, 2005).
• 8
  Note the strangeness of a literal translation ‘[…] in their own selves’.
• 9
  Note that the change of contained/container is one of the most frequent cases of semantic change by metonymy. See Luján (2010b).
• 10
  For an excellent study of the relation between ātmán- and prāṇá- , especially in later literature, see Zysk (1993).
• 11
  Note that Willman-Grabowska (1931) maintains that the idea of the ‘soul or inner principle’ or ‘oneself/ the self’ is more frequent that that of ‘vital breath’ in the AV.
• 12
  Péṣṭraṃ probably makes reference not only to the bone structure by to a section of bone with the surrounding flesh due to its possible etymological relation with peṣ- ‘strain, grind’ (MAYRHOFER, 1986-1992, p. 70), who translates it as Fleischstück (literally, ‘piece of meat’).
• 13
  “Rūpā́ni and ātmā́nam are antithetical: rūpā́ni ‘the external form of things’; ātmā́nam ‘its own nature’. It is a determining characteristic in Vedic that the internal nature, the true nature, of any divinity or instrument of power must be understood in order to control its influence or power”.
• 14
  I am especially grateful for the comments of one of the peer reviewers regarding the presentation of the semantic maps. I have altered the original by inserting directional arrows (absent in the model by HASPELMATH, 2003) in order to make clear that semantic change occurs in a particular direction.
• 15
  It should be noted here that the map has a special nucleus for emphatic values (‘the person themselves’). This does not necessarily mean that there are passages in which this is the function, but rather a semantic nuance.
• 16
  Svayám functions here as an adverbial intensifier: emphasising that the action is performed by Viśvakarman’s own means.
• 17
  Yaj- is habitually translated as ‘to sacrifice’ due to its relation with yajñá- ‘sacrifice’. However, there is a crucial difference between the behaviour of yaj- in Vedic and ‘sacrifice’ in Spanish: in the clause of the Spanish verb, the only second argument possible is that of the object of sacrifice, the oblation in Vedic, in addition to this construction, yaj- permits a second argument as the beneficiary of the sacrifice (BAUM, 2010). In these cases, the term is translated as ‘to worship’.
• 18
  Literally: ‘creating a friend for your own being’.
• 19
  In some way, in the spirit of the now classic proposal by Trier, beyond any criticism that linguistic historiography may make.

Publication Dates

• Publication in this collection
  May-Aug 2015

History

• Received
  Apr 2014
• Accepted
  Aug 2014