Verónica Orqueda: Reflexivity in Vedic

This monograph is a thoroughly revised version of the author’s doctoral dissertation, defended at the Universidad Complutense de Madrid in 2013. The book is a welcome contribution to Vedic and Old Indo-Aryan linguistics and in many respects represents a desideratum, since there exist no previous monographic treatments of reflexive strategies in Vedic (cf., however, Gonda 1979). The empirical base of the book is mainly drawn from the Rigveda but some material from the Atharvaveda has also been included, primarily from the Śaunakīya recension but also sporadically from the Paippalāda recension. The thematic scope of the book includes three constructions that serve as reflexive strategies in the Rigveda, the middle voice, the noun tanū́- ‘body’ and the noun ātmán-/tmán- ‘breath etc.’, as well as the possessive adjective svá- and the intensifier svayám ‘self’. There are eight chapters in the book, which also incorporates a list of figures and tables, a list of abbreviations, a survey of languages and texts, a brief explanation of how the cited passages have been selected, translated, and glossed, a bibliography and an index locorum. The first three chapters are introductory, the next four are thematically organized, and the last chapter contains the author’s final observations. The expressed aim of the author is “to gain new insights into the high complexity of the whole system [of reflexive strategies in Vedic]” rather than “to offer a concise description of Vedic reflexives” (p. 1).

Chapter 1 “Introduction” (pp. 1–14) is a brief introduction to the study of reflexives and reflexive strategies in general. It includes a section on the ancient Indian grammarians’ analyses of reflexivity (Section 1.1.1), a section on early western approaches to reflexivity in Vedic (Section 1.1.2) and a discussion of the analysis of reflexivity in the Government and Binding framework (Section 1.1.3), in cognitive frameworks (Section 1.1.4), in Functional Typology (Section 1.1.5) and in Construction Grammar and Radical Construction Grammar (Section 1.1.6). The chapter ends with a section summarizing the most recent contributions to the study of reflexivity in Vedic (Section 1.2). The main aim of the chapter is to discuss “very briefly the contributions and limitations of prior approaches to reflexivity in different languages.” The chapter contains several useful observations on reflexivity in general and has mostly found a nice balance as regards the level of detail in the discussion of the various topics. This has the advantage of providing the reader with some of the necessary background for following the analyses in the following chapters.

However, while it is quite understandable that the author has chosen to keep the introductory chapter relatively brief, one might object that some of the sections are somewhat less instructive than they could have been. One example of this is the section on how Indian grammarians dealt with reflexivity. Here, the reader learns that there is a rich indigenous grammatical tradition in India, which does not only include Pāṇini “but also earlier works śikṣās and prātiśākhyas” (p. 2). Somewhat surprisingly, however, the author fails to note that the bulk of works within this grammatical tradition are younger than Pāṇini (cf. e.g., Raster 2015 for recent discussion). It is noted (p. 2) that the pertinent sūtras in the Aṣṭādhyāyī “do not offer much of information regarding reflexivity as differentiated from other middle domains”, a claim that would seem to imply that Pāṇini does not distinguish reflexivity from more general middle voice semantics. The reader is thus left wondering whether the distinction between reflexive and other types of middle semantics was unknown to the Indian tradition. This issue might have been clarified by taking later works and commentaries into consideration.

The section on the “early western approaches”, that is, works dealing with reflexivity in Vedic (and Classical) Sanskrit from the second half of the 19th century until the 1970s is quite informative and notes that an important contribution made during this period was that nominal reflexive strategies were included as a topic of study. Both these sections are relevant for the topic of the book, but it is unclear why the author has chosen to present works dealing with reflexivity in Vedic from Mendoza Tuñón (1984) onwards in a separate section (Section 1.2), after a discussion of different theoretical approaches. This gives the impression that there has been some significant change in perspective within this field of research between 1979 and 1984 but there is no indication that this in fact is the case. As regards the sections on how different theoretical frameworks analyze reflexivity (Sections 1.1.3–1.1.6), they provide a convenient background for the development of the theoretical and methodological framework in Chapter 2 but suffer from a slightly superficial discussion. Taken together, the introductory chapter shows some minor weaknesses but clearly shows that the author has a good grip of the research tradition and different theoretical frameworks. Even if sometimes a bit too terse, the chapter provides the reader with sufficient background knowledge to follow the discussion and analyses in the later chapters.

The main aim of Chapter 2 “Theoretical bases and problematic issues regarding reflexivity” (pp. 15–36) is to provide “a concise description of the most problematic aspects associated with the study of reflexives, as well as systematizing the theoretical framework and methodological approach” (p. 15) followed in the book. The first section of the chapter provides a brief discussion of the semasiological and onomasiological approaches to reflexivity, concluding that the latter type of approach is most suitable for the object under study. The following section introduces direct and indirect reflexivity, noting that there is variation across languages as to the extent to which different types of indirect reflexivity are formally distinguished from each other and from direct reflexivity. We learn that Vedic expresses direct reflexivity and co-reference with a beneficiary – that is, self-benefactive reflexivity – through middle voice endings; whereas constellations where there is co-reference between the first argument and a recipient characteristically involve nominal markers (Section 2.1.1). Reflexive possession (Section 2.1.2) can be expressed by the possessive adjective svá- ‘own’, typically in combination with middle voice verb forms. The discussion of logophoricity (Section 2.1.3) notes that Latin can use reflexive pronouns with logophoric function. The author notes that “logophoric expressions may have an effect on valency, since they are closely connected to causatives” (p. 19). Unfortunately, this potentially controversial claim is not further substantiated, and it is not clear (at least not to this reader) in what ways these functions could be related. For the purposes of the present book, however, this is a rather minor issue, since Vedic does not employ reflexive strategies to express logophoricity. In many languages, including Vedic, there are polyfunctional markers with reflexive and intensifier readings, as is the case with tanū́- (e.g., Kulikov 2007). It is also noted (p. 21) that middle voice markers, on the other hand, can convey both reflexive and reciprocal functions in Vedic but it might have been emphasized that these two interpretations are characteristically associated with different classes of verbs, although some overlap might be expected.

Section 2.3 deals with the relationship between reflexivity and other, related domains, notably middle semantics (Section 2.3.1), transitivity (Section 2.3.2) and causativity (Section 2.3.3). This part of the chapter is overall clearly written and gives a convenient and concise overview of the most important views found in the pertinent research literature. Section 2.4 provides an outline of the theoretical and methodological framework on which the analyses presented in the book are based. The work has an emphatically functional-cognitive perspective, and one must agree with the author that this perspective offers a relatively clear way of approaching the problem under scrutiny (p. 31). The author notes that this implies the acceptance of polyfunctional markers and the co-existence of different markers with the same function. This approach is appealing, not least because it allows for capturing the potential ambiguity or vagueness of a marker such as tanū́-, which both represents a noun with the meaning ‘body’ and is employed as a reflexive strategy in Vedic. In most cases, the exact interpretation is not sufficiently clear to exclude one or the other meaning, a fact that finds a straightforward solution under the assumption that we are dealing with an ambiguous or vague word, as assumed in this book. One might object that an approach along these lines merely serves to mask our inability to arrive at a precise understanding of the Vedic texts. However, by carefully combining insights from grammaticalization theory, typology, and cognitive linguistics and employing the semantic map model, the author develops a model of grammar that has an explanatory scope well beyond the material analyzed in this study, thus showing that this approach is a fruitful way of approaching ambiguity and/or vagueness across languages.

Chapter 3 “The Vedic language and its texts” (pp. 37–47) gives some philological background, presenting a brief characterization of Vedic texts and a simplified chronological framework of Vedic (Section 3.1), a concise overview of its most pertinent morphosyntactic properties (Section 3.2), and a survey of its voice system (Section 3.3). Somewhat oddly, Vedic is defined as the language of the four principal saṃhitās and there is no mention of other (late middle and late) Vedic texts, notably the Brāhmaṇas and older Upaniṣads. Moreover, the author fails to note that the Vedic corpus includes both mantra and prose portions. Although she notes (p. 37, fn. 1) that she only includes texts relevant for the investigation, the book is published in a venue with a rather broad readership and is not primarily aimed at experts on Indo-Aryan, so one might expect a somewhat more general presentation of basic philological facts. One also wonders why the Role and Reference Grammar macroroles Actor and Undergoer suddenly appear in the discussion of the Vedic voice system, since the equivalent (?) terms employed in Chapter 2 were proto-agent and proto-patient as per Dowty (1991). However, taken together the chapter provides a good introductory presentation of some basic philological facts.

Chapter 4 “Middle voice” (pp. 48–127) is the first of four thematic chapters that form the kernel of the work under review. It deals with the middle voice, with particular reference to its use as a reflexive marker. The analysis is based on a comparison of middle forms from a sample of 27 verbs, which according to the author corresponds to ca. 10% of the verb roots given in McDonell’s (1916) verb appendix. The verbs have been selected and categorized based on the following criteria: 1) their relative importance in the Rigvedic corpus, defined in terms of their meaning and/or token frequency; 2) their pertinence within the other chapters of the study; 3) whether they form equivalence sets with verbs figuring in cross-linguistic studies dealing with reflexivity and/or transitivity. It makes perfect sense to limit a study of this kind to a subset of the existing verbs in the language under scrutiny; however, the above criteria are problematic in various respects.

What is perhaps the most serious problem is that they are partially formulated in a way that obscures the principles for selection. For example, what exactly does it mean that a verb has a meaning that is particularly important in the Vedic corpus? Given that the author explicitly notes that “prospective additions will always be welcome to challenge these findings” (p. 48), one would expect a higher degree of transparency as regards selection of predicates to facilitate replication and/or expansion of the analysis. The verbs are organized in 13 classes, but it is not always clear why a given verb is included in a given verb class. Why, for instance, is a verb such as añj- ‘anoint’ classified as a verb of putting rather than a verb of grooming, where we find a verb such as peś- ‘adorn’? Moreover, in other cases two verb classes seem to comprise partially overlapping sets of verbs. This appears to be the case with Verbs of Harm (Section 4.2.5) which includes the verb han- ‘smite, kill’ that might equally well be classified in the Verbs of Change of State group (Section 4.2.9). The class of Verbs of Engendering (Section 4.2.3) only has one member, jan ⁱ - ‘generate, be born’, so that one may entertain some doubts as to the generality of observations regarding verbs of this type. Another problem arises with a verb like bhan- ‘speak’ (p. 110), which the author notes is meagerly attested, a fact that is surprising, given that one of the criteria for including verbs in the sample in the first place was their relative importance (“relevance”) in the Vedic corpus in terms of meaning and token frequency. A final point concerns the verb han- ‘smite, kill’ which according to the author does not show any examples of reflexive use of the middle (p. 78). She remarks that “this may be due to different reasons” but adds that there are no theoretical restraints that would preclude the possibility of reflexive middle voice with this verb. Unfortunately, she does not specify what reasons could explain this. While these points certainly do not weaken the general validity of the claims made in this work, they illustrate the importance of using clear-cut criteria for classifying verbs or other types of word in studies like the present.

Chapter 5 “SVÁ- and SVAYÁM” (pp. 128–144) examines the use of the possessive adjective svá- and the intensifier svayám ‘self’, which makes sense, since they are clearly formally related. As noted in Section 5.1, these forms ultimately derive from a Proto-Indo-European form *s(e)u̯e-, which has given rise to reflexive markers in several Indo-European languages. Section 5.2 discusses their etymology and meanings, noting that svá- is generally taken to have a reflexive meaning, being able to have any person as antecedent. The author also mentions that svá- can be used as a collective noun denoting “a set of belongings, a group of human belongings, as well as the persons themselves” (p. 129). It is also used as an emphatic or intensifying element, which reflects an important difference between svá- and middle verb forms, which do not have an emphatic or intensifying use. svayám, on the other hand, bears a formal resemblance to svá-, and is commonly analyzed as reflecting the Indo-Iranian dative singular form *svai- with the extension -ám, analogically influenced by the nominative forms of pronouns like ahám ‘I’, t(u)vám ‘you [2sg]’, or ayám ‘this here’. This derivation clearly is perfectly reasonable since svayám is an intensifier which occasionally occurs with (expressed or unexpressed) nominative pronouns and nouns (e.g., RV I 122.9, I 129.6, V 55.2 etc.). However, given the general tendency that intensifiers develop into reflexives, it is surprising that the author seemingly derives the intensifier svayám from an indeclinable reflexive particle *sué (p. 130).

An advantage of this section is that it provides a good overview of the various available analyses of these forms but the discussion is not always easy to follow and one might have wished for a somewhat more explicit conclusion. Section 5.3 provides a survey of the spectrum of uses of svá- which is attested 183 times in the corpus (183 in the RV, 50 in the AVŚ), noting that “the possessive reflexive nature of some hundred passages cannot be ruled out” (p. 132). It is also observed that the possessive reflexive use of svá- shows complementary distribution with “non-emphatic (reflexive) possessive strategies: personal (táva) and demonstrative pronouns (like ásya)” (p. 133). In other cases, svá- is used with a non-reflexive possessive meaning, showing that it is not restricted to the reflexive function. Furthermore, the author notes that “given that the presence of svá- is not necessary to indicate a reflexive possessive where there are other resources, the reflexive possession can be seen as secondary” (p. 135). I found this argument difficult to follow, and it might have helped clarify matters if the text provided a table or another graphic representation to summarize these observations. It should also be pointed out that svá- on occasion modifies tanū́- ‘body, self’,^[1] even if the author claims that “they are not always part of the same NP” (p. 135–136). I am not quite sure to interpret this, but I assume that it reflects the fact that the two constituents can be discontinuous. Here, a couple of illustrative examples would have helped clarify what is intended.

Section 5.3 deals with the intensifier svayám, of which 29 examples are found in the corpus.^[2] The author notes that svayám characteristically occurs with middle verb forms, sometimes in combination with tanū́-. It is often used in self-benefactive contexts but according to the author, this is a consequence of its intensifying function. Another intriguing observation concerns the fact that svayám is compatible with inalienable and alienable possession (p. 138–139, with examples [20] to [22]). However, the examples are somewhat confusing (ánnam ‘food’ in (20) is classified as inalienable possession, mahitvám in (22) as alienable possession). Later, commenting on example (25), it is stated that “svayám emphasizes only the autonomy of yajasva, although one cannot rule out that the merge of all these elements (svayám, tanvám, a middle verb form) in the same clause may have an influence on the interpretation” (p. 140). Again, the argument is difficult to follow. Why shouldn’t all the elements in question have an influence on the interpretation of the clause? A similar problem concerns the discussion of example (29) on p. 141, where we learn that “[g]iven that this is an early hymn, the intensification of the co-reference is highly expected, while that would not be the case for more recent hymns. Unfortunately, given that this is the only passage with these characteristics, it is difficult to come to a well-founded conclusion.” First, why is intensification of co-reference highly expected in early hymns but not in more recent ones? Second, if this is the only example with the pertinent characteristics and it is difficult to reach a conclusion, on what basis has the prior observation been made? This is entirely ad hoc.

On p. 142, it is noted that svayám tends to be preferred to tanū́- in intensifying function and that this is the case in older hymns, since tanū́- allegedly intensifies contexts inviting a body-term interpretation. While this seems to be a reasonable observation per se, it has not been demonstrated that it is the case anywhere in the previous text and hence is stipulative. It should also be pointed out that the intensifier svayám is only found in Indo-Aryan, whereas tanū́- has an analogous use in Iranian, notably Avestan. Thus, it would be reasonable to regard the intensifying use of tanū́- as an Indo-Iranian inheritance and therefore older than the use of svayám. Under this analysis, one would expect tanū́- to be the preferred intensifier in older stages of Vedic, being subsequently substituted by svayám. This expectation seems to be borne out to some extent, since svayám, unlike tanū́-, remains in use in this function after the Rigveda. Finally, the author raises the question whether the “ambiguity between intensification and reflexivity in contexts where svayám appears, may be caused by the similarity between this ending and that of the great majority of masculine and neuter nouns in the accusative case” (p. 143) and concludes that “no significant data have been found to confirm this possibility; there are many other forms of the accusative, which would also have to be taken into account and furthermore, other types of parts-of-speech may have a similar ending” (p. 143). Here too, the argument is difficult to follow. While this chapter presents an illuminating outline of the functional range of svá- and svayám, some problems remain open for future research.

Chapter 6 “TANŪ́-” (pp. 145–184), deals with the word tanū́- ‘body, self’, which in some cases has a reflexive meaning. Its use as a common noun is considerably more frequent than its reflexive use, and it is often difficult to determine with certainty which interpretation is predominant. As previously noted, this problem is to some extent resolved in this work by assuming that such categories are polyfunctional and inherently ambiguous, resulting from the grammaticalization of a body-part noun via an intensifier to a reflexive marker. Nevertheless, Section 6.3 deals with the different non-reflexive nuances of meaning associated with tanū́-, notably ‘body’ (Section 6.3.1), ‘person’ (Section 6.3.2), ‘shape’ (Section 6.3.3) and ‘intensifier’ (Section 6.3.4), while Section 6.4 explores its use as a reflexive marker. These discussions show that tanū́- competes with ātmán-/tmán- ‘breath, life, person’ in the second and fourth meanings/functions and with svayám in the fourth function. It should be noted, however, that the use of ātmán- with the meaning ‘person’ first surfaces in the AV. The cross-linguistic comparison on p. 154 intended to illustrate that nouns with the meaning ‘person’ is common in many languages only includes examples from Spanish and English, thus leaving something to be desired. In the discussion of the interaction between tanū́- and different verb classes, it is noted that in cases where it is used with verbs of grooming, there are no cases of disjunctive reference (p. 169), a fact possibly suggestive of inherent reflexivity.

As regards the verb śumbh- ‘beautify, embellish’, the author states that the verb “despite being an event performed on one’s own body, needs a heavier marker to indicate that it is an event performed on oneself” and the behavior of this verb is compared to prototypical grooming verbs like English shave which do not select a reflexive in object function. However, this comparison is somewhat dubious, not least because the English equivalent of the Vedic verb demands a reflexive pronoun in cases of co-reference between first and second argument, cf. The girls embellished themselves. This point illustrates the problematic nature of the general classification of verbs in this work, such as verbs of grooming, which do not necessarily constitute a group of verbs with corresponding morphosyntactic behavior. On the other hand, the discussion on pp. 170–171 provides a good analysis of the development from [corporeal body] to [reflexive] with verbs of grooming (cf. also the semantic map on p. 158).

Chapter 7 “ĀTMÁN- and TMÁN-” (pp. 185–206) explores the use of ātmán-/tmán- ‘breath, vital principle, self’ in the RV and AV. Even though these two forms originally represented different stem forms of the same paradigm, the strong stem ātmán- has been generalized as a noun stem in the RV, while the weak stem tmán- seems to be independent of the former. There are apparently no examples of the noun ātmán- in direct reflexive function in the RV but it does show this function in the AV. The form tmán-, on the other hand, occurs once in direct reflexive function in the RV but is often used with an apparent intensifying or indirect reflexive use. This form is virtually obsolete in the AV, where there are only four examples. Interestingly, reflexive ātmán- predominantly appears with active forms of the verb, unlike tanū́-, which almost exclusively occurs with middle forms. The author notes (p. 204) that ātmán-/tmán- has the appearance of representing more of a grammatical device than tanū́-. Under her analysis, this is reflected in the fact that it shows reduction in case inventory, and that it is independent of other strategies, e.g., middle voice and the appearance with svá-.^[3]

Chapter 8 “Final observations” (pp. 207–218) conveniently summarizes the main findings of the previous chapters. After a short introduction (Section 8.1), Section 8.2 contains a subsection on each of the possible reflexive strategies in the stage(s) of Vedic that fall under the scope of the present work, culminating in a semantic map illustrating the author’s analysis of the relationship between the various strategies. Section 8.3 discusses the diachronic development of the system of reflexive marking in Vedic, while Section 8.4 outlines the work’s contributions to the typological study of reflexives. Here, we learn that the middle voice is a more productive strategy for expressing reflexivity in the earliest layers of the Rigveda, a trait probably inherited from Proto-Indo-European. Interestingly, approximately half of the attestations of tanū́- occur in books I and X, which are generally regarded as slightly later than the rest of the Rigvedic corpus. Although the author is probably right when claiming that “the number of possible reflexive cases with tanū́- at this stage is scarce and that there is no complete grammaticalization yet” (p. 214), she fails to mention that tanū́- is occasionally employed as a reflexive strategy in Avestan, where it has an equally broad range of meanings as in Vedic. This fact might be taken to suggest that the strategy of employing tanū́- as a reflexive strategy is of Indo-Iranian age, indicating that the prehistory of the Vedic reflexive system is somewhat more complex than the author suggests.

The monograph concludes with a Bibliography and the “Index of cited texts”. Unfortunately, the passages cited from the Paippalāda recension of the Atharvaveda are not given among the text passages cited in the index locorum. Finally, a general index of key topics, authors cited etc. would have been very helpful for the reader.

Although this recension has had a rather critical perspective, I wish to emphasize that this book is a welcome contribution to the field of Vedic linguistics. The author should be commended for insightful discussion of difficult data and intriguing analyses of sometimes difficult and obscure material. Being somewhat of a pioneering work, it may be expected to spur other, similar studies in the future.

Minor points:

p. 73. “…does not admit -á- forms” > “…does not admit -yá- forms”

p. 81. “Show verbs (avís kr̥-) [> āvís kr̥-]” – only one predicate, why “show verbs” and not simply ‘disclose, reveal’? cf. p. 78 ‘conceal (guh-)’

p. 91, ex. (115): careme should be caratas and glossed as 3du.prs.m

Exx. (45) and (49) on pp. 105–106 are identical (AV 13.2.10) but are intended to illustrate two different readings (reflexive possessive vs. self-benefactive).

p. 151, l. 3: competence > competition

p. 200, ex. (36): g-acc.sg > woven.at.home:acc.sg