On the function of saṁhitā in the Saṁhitā Upaniṣad

The Saṁhitā Upaniṣad [SU] is a little-known Vedic text that presents ‘typical’ Upaniṣadic teachings on the truth of identity alongside seemingly out-of-place descriptions of rites used to protect oneself against enemies and even against death. The difference between these contents is striking, but what it has to tell us about the SU’s main concerns is vulnerable to historical and text critical methods that rely on structure, style, and linguistic archaism to divide texts into discrete strata. What if the modern text critical practice of individually identifying and classifying textual contents obscures the use and meaning of the word saṁhitā in the SU? Is it possible that the SU’s diverse contents are intrinsically related? This article explores these questions through a close examination of a sequence of passages illustrating the contrast that has led previous scholars to see the SU as miscellaneous in character and lacking internal coherence. Through this examination, I identify a wider context for saṁhitā in the specific relationship the SU depicts between the person (puruṣa) and speech (vāc). I argue that the SU’s treatment of saṁhitā draws upon an understanding of recitation in the perspective of one’s vulnerability and the dynamics involved in developments of personhood. These findings allow the SU to emerge as an intriguing and coherent text that merits closer examination and establishes a promising approach for the study of the R̥gvedic Āraṇyakas.


Introduction 1
The Saṁhitā Upaniṣad (SU) of the Aitareya Āraṇyaka (AitĀ) is by all accounts a little-known Vedic text. Unlike its "sister" 2 text the Aitareya Upaniṣad (AitU), the SU has not been included in the majority of published studies and editions of the Upaniṣads Which represent them as a textual genre or division of the Veda. 3 The reason for this appears to be quite simple. The diversity of the SU's contents is ill-suited to modern text critical methods that rely on structure, style, and linguistic archaism to divide texts into discrete and distinctly identified strata. Approached from this angle, the SU is defined more by what is perceived as its inconsistency than by the concerns and literary traits that it shares with many other Upaniṣads. Yet, considering that the Upaniṣads have been described as an "amalgam" held together by "overlapping" concerns and strategies that "can be seen as loosely constituting a genre" (Lindquist, 2020, p. 2), we might ask what it is about the SU in particular that has made it worthy of exclusion from modern depictions of the Upaniṣads as a corpus.
Two considerations stand out. First is the sheer extent of the SU's heterogeneity. 4 More than anything else, it is arguably this feature of the text -the continual shifting from ruminations on the cosmos and the truth of identity to a preoccupation with malign omens and the personal use of imprecations -that seems to have drawn scholarly attention away from the possibility of an underlying consistency of focus. 5 A second consideration is the impact of previous interpretations of what the 1 All Sanskrit translations are my own unless indicated otherwise 2 Cohen (2018, p. 261).
3 I thank the anonymous reviewer who pointed out that Max Müller's (1879) translation of the SU is a significant exception to this trend; see Sacred Books of the East Vol. 1. Examples of publications which name the SU in relation to the AitU but otherwise omit it include Deussen 1980Deussen [1897, Olivelle 1998, andCohen 2018. 4 The major considerations at play are questions of if and how to reconcile differences in textual contents that on the surface suggest the lack of an intrinsic relationship between compositional sequences -a tendency that is often complicated by inherited notions about the irreconcilability of "magical" versus properly theological or revelatory forms of religious discourse. Although scholarly critiques of this interpretative stance are reasonably widespread, the practice of assuming the presence of an actual, qualitative dimension behind proposed distinctions in what is magical and what is religious has not yet shifted sufficiently to allow newly inclusive typologies to emerge. To cite Patton (2005, p. 18): "It is now fairly widely accepted that both scholars and theologians working from within a tradition use the term magic to delineate less properly "theological" forms of religious discourse. However, the critique should not stop there. The term also serves to cut off important social and exegetical continuities between a religious tradition and its so-called magical counterpart. It drives a wedge between forms of thought, which, from the tradition's eyes, may be integrally connected." 5 As Paul Deussen (1980 [1897], pp. 12-13) states in the introduction to his translation of the AitĀ, "Āraṇyakam 3 [= the SU] deals with the secret meaning (Upaniṣad) of the letters and the connections among the letters which, as it occurs also in a different way with the imparting to the student of knowledge held as secret (Br̥ h. Up. 6.4), is joined to the Upaniṣad; but according to its contents it is so heterogenous [sic] that here we can leave it out of consideration."A related factor behind the SU's omission from the corpus of Upaniṣads may, tentatively, be the influence of the commentator Śaṅkara's example. I have not been able to trace an explicit statement of this regarding the SU, but we might consider Rajendralal Mitra's (1876, p. 18) comment that while all of AitĀ 2 and 3 are considered an Upaniṣad by Śaṅkara, the latter "separated the last four chapters of the second Book [AitĀ 2.4-6 = AitU], as Upanishad par excellence." Keith (1909, p. 17) uses the same wording to describe the AitU. Müller (1879, p. xcvii) places similar emphasis on the (canonical?) position of the AitU over the other Upaniṣads of the AitĀ: "we must SU means when it speaks of saṁhitā. According to the seminal works of Max Müller (1869Müller ( , 1879, saṁhitā refers to the sacred text of the R̥ gveda (R ̥ V) "in which all letters are closely joined" 6 or have undergone euphonic combination (sandhi) resulting in the specific recitation type known as saṁhitāpāṭhā. 7 This interpretation was followed by A. B. Keith in his edition and translation of the AitĀ in which he describes the SU as an exposition of "the mystical meaning of the various forms of the text of the [R ̥ V] Saṁhitā, the nirbhuja, pratr̥ ṇṇa and ubhayamantareṇa, and of the vowels, semi-vowels, and consonants" (1909, p. 17). 8 While it is certainly the case that SU demonstrates considerable concern with the deeper significance of the recitation and pronunciation of the R ̥ V, this interpretation of saṁhitā cannot accommodate the majority of the SU's contents -its bandhus, 9 rituals, descriptions of impending death, and cosmological accounts. Indeed, it is possible that the very delimiting effect of such an interpretation of saṁhitā can be seen in Keith's assessment that aside from those passages that directly speak of nirbhuja, the rest of the SU's contents contain "little of philosophical interest" (1909, p.44). 10 While a historical link between the way that saṁhitā is defined and the way that the SU has challenged our understanding is apparent, there is a further factor that has influenced the tendency to overlook the possibility of the SU's coherence. One result of the SU's omission from the standard corpus of Upaniṣads is that it has instead been included in scholarly work on the AitĀ. However, this has not led to an improved understanding of the relationship between the SU's contents and its proposed focus on saṁhitā for the reason that the Āraṇyakas as a genre are broadly perceived as containing miscellaneous "strange" materials and functioning as a "catch-all" for later Vedic teachings (Witzel and Jamison 1992, pp. 5, 12). By contrast, the difficulties surrounding the study of the SU are emphasized by the fact that while neither the distinguish between the Aitareya-upanishad, properly so-called…and the Mahaitareya-upaniṣad, also called by a more general name [Bahvr̥ ca]-upanishad, which comprises the whole of the second and third [Āraṇyakas]." We may also compare Deussen's (1980Deussen's ( [1897, 12) comment on his decision to exclude AitĀ 2.1-3, the upaniṣad directly preceding the AitU, from his translation: "…the Upaniṣadic thought here is so little clear and is so much mixed with what does not belong to it, that we prefer to follow the example of Śaṅkara and exclude Ait. Ār. 2.1-3 from its inclusion in the Upaniṣad, all the more so as the precious thoughts in the Kauṣītaki Upaniṣad recur, partly verbatim." 6 Müller (1879, p. 246, n. 2). This interpretation of what saṁhitā intends or implies as the focus of the SU was also influential in the works of Oldenberg (1888, 1892) and Macdonell (1900), both of whom cite the SU when discussing the chronology of the R ̥ V's formation as a canon. 7 The SU refers to this as nirbhuja; see SU 3.1.3 and 3.1.5.
8 See also Keith's discussion of the parallel SU in the Śāṅkhāyana Āraṇyaka (1908, p. xi). 9 To cite Smith (2006, p. 211), a bandhu is a "linkage of one phenomenon, entity, being, notion, or concept with another, based on phonological similarities, numerological equivalences, or other formal principles" that is typically generated at a syntactic level through the utilization of simple nominal apposition with or without the emphatic particle vai ("X vai Y"). While the extreme syntactic brevity of bandhus does not belie their metaphysical depth, their lack of expository detail has left scholars in conflicting opinion over how these linkages should be understood. 10 It is of note that Mitra does not describe the SU from the perspective of a definition of saṁhitā but instead asserts that the three Upaniṣads of the AitĀ contain teachings intended for different audiences. According to his scheme (1876, p. 18) the SU is aimed at the edification of "those who have no idea of emancipation, and long for progeny, cattle and the like: these are the lowest and most grovelling [class of men]." SU nor the AitU focuses on the Mahāvrata ritual that forms the main subject-matter of rest of the AitĀ's contents, the AitU has been distinguished as the AitĀ's upaniṣad "par excellence" (Keith 1909, p. 17) while the SU's contribution to the study of the Āraṇyakas -or of the Upaniṣads -has yet to be addressed. Again, it appears that text critical methods aimed at the classification of texts based on distinctions of style, structure, or commonality of content struggle to identify underlying points of convergence, or resort to characterising texts and genres by their mixed and miscellaneous status.
These observations linger in the background of the present article. However, rather than aim to produce an analysis of the SU in entirety, my focus is on the preliminary task of asking how a broader definition of saṁhitā can develop our understanding of the SU's scope, and perhaps even its purpose. Indeed, does saṁhitā possess a scope beyond a simple variety of R ̥ gvedic recitation? Are there other circumstances that it can be applied to or which it informs? Are there correlating factors that come into play when saṁhitā is viewed against the wider concerns and themes articulated in the SU? If priority is placed with the traditionally transmitted form of the SU, and I contend that there is no overriding reason for it not to be, then we are pushed to consider possibilities beyond the disjunctures perceived by scholars over the past two centuries -in particular, the possibility that there is in fact a relationship between the SU's conception of saṁhitā and the variegation of its contents.

The SU: Approach, Textual Selection, Preliminary Issues
If it is accepted that previous definitions of saṁhitā as a style of recitation are potentially narrow, precluding the consideration of apparently different kinds of textual content and influencing the perception of miscellany in the SU, then its re-examination requires an approach that preserves differences, seeing an opportunity to identify underlying points of convergence in the tension between contrasting materials. With such criteria in mind, the major part of this article focusses on a sequence of three passages (AitĀ 3.2.2-4) occurring in the SU's second adhyāya. Although AitĀ 3.1 contains many materials for the examination of saṁhitā, it is in the second adhyāya that we encounter suggestive evidence of a relationship between saṁhitā and the individual person (puruṣa), thus opening the our examination of saṁhitā to questions of how a characteristic of recitation might have been viewed as personally relevant in the world of the text.
These passages contain some of the SU's most challenging contents, presenting a series of bandhus on saṁhitā alongside a lengthy description of symptoms associated with impending death. However, while these passages illustrate the very heterogeneity Keith (1909, p. 251 n. 7) took as a mark of inconsistency in the ordering of the SU's second adhyāya, I argue they form a coherent sequence but with demonstrable shifts in their respective orientations towards saṁhitā. Secondly, I explore the possibility that these multiple perspectives are related by a set of discernible background concerns regarding the identity of the puruṣa, its establishment and loss, and the role of sacred speech (vāc) in the form of recitation in disclosing and maintaining the connections upon which personal attainments depend.

Vāc and the Puruṣa
Like many other Vedic texts, the AitĀ contains descriptions (AitĀ 2.1.4-5 and 2.4.1-2 11 ) of the puruṣa as consisting of numerous deities (devatā) that each contribute their own unique characteristics to a person's physiological operation. It is from this collocation of influences that the identity of the puruṣa emerges as a 'combination identity' that can be divided into constituent factors. According to AitĀ 2.1.4-5 and 2.4.1-2, the puruṣa is entered by devatā -sight, hearing, mind, speech, and breath 12 -during its creation, causing it to arise as a functioning whole. Yet, following this, the texts describe the devatā abandoning the puruṣa while the puruṣa is livingrendering it deaf, dumb, mute, and so on in turns -before re-entering it. I note that these accounts depict the devatā as lively and capable of exerting their own volition. Furthermore, the devatā are described as each possessing a network of identifications and relationships beyond their respective physiological functions, 13 and the traces of these wider relationships are introduced into the puruṣa along with the movements of the devatā. 14 The significance of these unions 15 of identity within the devatā and the repercussion of their transactions into the puruṣa are a repeated topic. AitĀ 2.1.5 indicates that what the devatā have not shared is not embedded (prahita) in the self, but that the unions that have penetrated (niviṣṭa) the body can be known. 16 AitĀ 2.1.7 correlates the devatās with specific "powers" (vibhūti) contained in the puruṣa; 17 and in AitĀ 2.1.8 the devatās are described as the "hiding place" (giri) of the gods and of brahman in an elaboration on the text's eponymous teacher, Mahidāsa Aitareya's 11 AitĀ 2.4.1-2 = AitU 1.1-2.
12 These five devatā are listed at AitĀ 2.1.4-5 and again at 2.1.7-8. AitĀ 2.4.1-2 (= AitU 1.1-2) cites a longer list, which includes the former plus scent, hairs, and semen, and replaces prāṇa (breath, outward breath) with apāna (breath, downward breath). 13 Indeed, these are the microcosmic-macrocosmic relationships that are articulated using bandhus in a great many Vedic texts. The relationships listed in AitĀ 2.1.5 are fire-speech, sun-sight, moon-mind, and directions-hearing; a slight distinction is that it identifies prāṇa with the day and introduces apāna in identification with the night (ahar eva prāṇo rātrir apānāḥ). AitĀ 2.4.2 (= AitU 1.2) gives a longer list that extends to each devatā's physical support (āyatana) in the body: fire-speech-mouth, air-scent-nostrils, sun-sight-eyes, directions-hearing-ears, plants/trees-hairs-skin, moon-mind-heart, death-apāna-navel, and waters-semen-penis. Prāṇa is re-introduced at a later stage of this cosmogony (AitĀ 2.4.3 = AitU 1.3). 14 For an extended discussion of this phenomenon, including its depiction in Vedic texts, see Frederick M. Smith The Self Possessed (2006). I note that Smith draws upon earlier work by the anthropologist McKim Marriot; the following of Marriott's observations is particularly salient in the context of the puruṣa: "it appears that persons are generally thought by South Asians to be "dividual" or divisible. To exist, dividual persons absorb heterogenous material influences… Persons also cannot help exchanging certain other coded influences that are thought of as subtler, but still substantial and powerful forms, such as perceived words, ideas, appearances, and so forth" (Marriott 1976, p. 111). 15 The SU employs alternates between the terms saṁhitā and sandhi to express the sense of "union" between two or more individually identified referents (saṁhitā = AitĀ 3.1.1, 5-6; sandhi = AitĀ 3.1.2, 3.2.1). By contrast, AitĀ 2.1.5 employs the term saṁyoga in the context of the devatā. The analysis of these employments is desirable but beyond the scope of the present discussion. 16  statement that he knew himself as extending to the gods and the gods to him. 18 In the SU, the same devatā are identified in the repeated statement that the self is composed of sight, hearing, the mind, and speech, to which is added the meters. 19 Despite the fact that such accounts are well-known, they raise a point that has not received much attention -namely, their ability to nuance our reading of texts like the SU because they tell us something important about the way their human subject, the recipient of their teachings, was thought to have a porous and relational identity with the external world. This provides a conceptual context that allows us to think laterally about the way that adjacent, yet strikingly diverse passages might share common assumptions. Equally significant, these accounts suggest that teachings about speech are teachings about a constituent in the formation of personal identity, and they provide a scaffold for teachings that associate the extended identification of speech -with the consonants, syllables, metres, verses, sandhi, and indeed with recitation -with modifications to the identity of the puruṣa. 20 Although the descriptions mentioned above of the puruṣa's possession by devatā are constrained to the two Upaniṣads that directly precede the SU, their interests and the dynamics they imply are echoed by the SU; they are also resonant with the dual concern with speech and breath (prāṇa) that appears throughout the AitĀ, the SU included. It is viable that understanding saṁhitā in the SU requires that we bear in mind the relationship between speech and the puruṣa, seeing it as an ever-present consideration.

The transforming identity
The movements of the devatā -their entering and exiting of the puruṣa -raises a pair of further considerations relating to the puruṣa and the formation of personal identity. These have previously been discussed by Smith (2006, pp. 18-23) but bear reiteration in relation to the SU.
The first consideration is that the puruṣa's openness to external elements and influences renders it vulnerable: the achievements that result from true knowledge can be compromised and undone by influences originating outside of the puruṣa's bodily boundaries. This concern with potentially destructive influences and the 'boundary crossing' between external factors and internal effects are widespread in the Vedic  Sound and Sign (2007) which examines grammar -a trace or "substance code" (Marriott, 1976, p. 111;Smith 2006, pp. 20-21, 74-75, 187, 203, 214) contained in vāc -in its familiar mode as a set of rules governing the use of a language from the perspective of grammar's regular appearance in the creation accounts of numerous religious traditions and its subsequent metaphysical value. See in particular their observation that grammar "presents a methodology much more textured and elastic than other conceptual models" particularly in its capacity to contain "semantic and trans-semantic pieces of reality" and allow for "structured transitions between disparate domains" (La Porta & Shulman, 2007, p. 6).
traditions -recall, for example, the numerous instances in Atharvan literature 21 -and in the SU it takes particular shape as an anxiety about imprecations or curses. 22 Interestingly, this concern is clearly demonstrated in AitĀ 3.1.3, one of the two passages that Keith, following Müller, cited in evidence of the interpretation that the SU's major focus is the metaphysical significance of the recitation types. According to AitĀ 3.1.3: Next come the sayings about the nirbhuja. 23 The nirbhuja has the earth as its abode; the pratr̥ ṇṇa has heaven as its abode; the ubhayamantareṇa has space as its abode. Now, if (someone) curses one who is pronouncing the nirbhuja, he should say "You have fallen from the two lower conditions." Now, if (someone) curses one who is pronouncing the pratr̥ ṇṇa, he should say "You have fallen from the two higher conditions." But there is no cursing the one who pronounces the ubhayamantareṇa. For when one produces the conjunction (sandhi), that is the nirbhuja-form, and when one utters the two syllables pure, that is the pratr̥ ṇṇa-form, (but) foremost indeed is the ubhayamantareṇa -it extends to both. One who desires proper food should pronounce the nirbhuja; one who desires heaven, the pratr̥ ṇṇa; one who desires both, the ubhayamantareṇa. Now, if another curses one who is pronouncing the nirbhuja, he should say "You are obstructing the deity (devatā), the earth. The deity, the earth, will ruin you." Now, if another curses one who is pronouncing the pratr̥ ṇṇa, he should say "You are obstructing the deity, heaven. The deity, heaven, will ruin you." Now if another curses one who is pronouncing the ubhayamantareṇa, he should say "You are obstructing space, the deity. The deity, space, will ruin you." Whatever he is saying or might say to the (other) one speaking is surely achieved. But one should never speak infelicitously to a brāhmaṇa. Only when in exceeding prosperity should one speak back to a brāhmaṇa. "Not even in exceeding prosperity should one speak back to a brāhmaṇa, brāhmaṇas must be honoured" -so has said Śūravīra Māṇḍūkeya. 21 See Marko Geslani's (2018) analysis of śānti rituals in Atharvan ritual. Similar dynamics of transfer across bodily boundaries are evidenced in the Upaniṣads. Examples include the rites of transfer (saṁpratti) in Br̥ hadāraṇyaka Upaniṣad 1.5.17-20, the transfer between father and son (saṁpradāna) in Kauṣītaki Upaniṣad [KṣU] 2.15, and the new and full moon rites described at KṣU 2.8-9. 22 AitĀ 3.1.4 is entirely dedicated to "imprecations" (anuvyāhāra). 23 AitĀ 3.1.3: athāto nirbhujapravādāḥ | pr̥ thivyāyatanaṁ nirbhujaṁ divyāyatanaṁ pratr̥ ṇṇam antarikṣāyatanam ubhayamantareṇa | atha yady enaṁ nirbhujaṁ bruvantam upavaded acyoṣṭāvarābhyāṁ sthānābhyām ity enaṁ brūyat | atha yady enaṁ pratr̥ ṇṇaṁ bruvantam upavaded acyoṣṭā uttarābhyāṁ sthānābhyām ity enaṁ brūyat | yas tv evobhayamantareṇāha tasya nasty upavādaḥ | yad dhi saṁdhiṁ vivartayati tan nirbhujasya rūpam that yac chuddhe akṣare abhivyāharati tat pratr̥ ṇṇasyāgra u evobhayamantareṇobhayaṁ vyāptaṁ bhavati | annādyakāmo nirbhujaṁ brūyāt svargakāmaḥ pratr̥ ṇṇam ubhayakāma ubhayamantareṇa | atha yady enaṁ nirbhujaṁ bruvantam para upavadet pr̥ thivīṁ devatāmāraḥ pr̥ thivī tvā devatā riṣyatīty enaṁ brūyāt|atha yady enaṁ pratr̥ ṇṇaṁ bruvantaṁ para upavaded divaṁ devatāmāro dyaus tvā devatā riṣyatīty enaṁ brūyāt | atha yady enaṁ ubhayamantareṇa bruvantaṁ para upavaded antarikṣaṁ devatāmāro 'ntarikṣaṁ tvā devatā riṣyatīty enaṁ brūyāt | yathā tu kathā ca bruvan vā bruvantaṁ vā brūyād abhyāśam eva yat tathā syāt | na tv evānyatkuśalād brāhmaṇaṁ brūyāt | atidyumna eva brāhmaṇaṁ brūyāt | nātidyumne cana brāhmaṇaṁ brūyān namo astu brāhmaṇebhya iti ha smāha śūravīro māṇḍūkeyaḥ || My sense is it can be argued that the focus of this passage is not on the "mystic meaning of the various forms of the [R ̥ V] Saṁhitā" (Keith, 1909, p. 17) -that is, of the saṁhitāpāṭha (viz., nirbhuja), padapāṭha (= pratr̥ ṇṇa), and kramapāṭha (= ubhayamantareṇa) types of recitation -but rather on the practical and personal implications of each. Although what stands out about this passage at first glance is its positioning of recognisably 'religious' Vedic speech -recitation and bandhusalongside varieties of speech that seem more superstitious than religious -curses and "infelicitous" (anyatkuśala) speech -both kinds draw upon the same network of extended identifications that each recitation-type contains. The difference is one of acquisition versus loss, and while the perspective on the puruṣa's openness to the impacts of vāc remains the same it informs a context in which there is an articulated need for the preservation of one's achievements lest they be pulled apart by someone else's speech act.
A second consideration raised by the relationship between the puruṣa and the devatā is that it is the puruṣa that provides the basis for transformations of personal identity. This is because it is the puruṣa that can and does change, and in which religious achievements take hold, as distinct from the idea of one's 'ultimate' identity (commonly referred to as ātman) which is transcendent of any individual person. This allows for a further distinction to be drawn between the archetypal puruṣa whose creation and relationship to the devatās is a repeated topic of the two Upaniṣads that comprise AitĀ 2 and the puruṣa on a personal level, that is, as the substrate for individual personhood or identity and its development.
Significantly, it is the latter that occupies the SU's focus. In fact, unlike the Upaniṣads in AitĀ 2 which pay extensive attention to the truth of brahman and its revelation as the ultimate identity of the puruṣa, 24 the SU's attention is dominated by its focus on the puruṣa's containment of tightly compacted relationships (saṁhata) and "unions" (Keith tr. saṁhitā, sandhi) between devatā. 25 These compacts and unions are presented as instruments of personal attainment. Thus, in AitĀ 3.1.1: sa eṣo 'śvarathaḥ praṣṭivāhano manovākprāṇasaṁhataḥ | sa ya evam etāṁ saṁhitāṁ veda saṁdhīyate prajayā paśubhir yaśasā brahmavarcasena svargeṇa loken sarvam āyur eti | This compact of mind, speech, and breath is a chariot with a team of yokehorses. One who knows thus this union is united with offspring, livestock, glory, the splendour of brahman, the world of heaven -he lives an entire life.
Bringing these considerations together, we see that saṁhitā has a range of employment that applies explicitly to personal identity and the means of developing it, that recitation is one in a variety of verbal activities the SU acknowledges as having personal impact, positive or negative, and that the coherent basis for all of this lies with a vision of the puruṣa as open and vulnerable but already embodying the macrocosmic relationships that bring regularity to the external world. With this in mind, I turn to the closer examination of the way these issues play out in AitĀ 3.2.2-4.

AitĀ 3.2.2-3.2.3: Kauṇṭharavya and Bādhva
The SU's second adhyāya commences with a sequence of three passages (AitĀ 3.2.1-3) presenting bandhus between speech, the self, and the sun or year. The initial two passages are closely related and appear repetitive, but upon analysis contain a progression of bandhus that establishes the backdrop to the third passage, AitĀ 3.2.3, allowing us to detect a purposeful development of focus leading to the discussion of death-symptoms in the SU's fourth passage. In this section, I will examine the second and third passages from the perspective of what I take to be a sequential relationship gradually building to the SU's paramount description of saṁhitā in the context of personhood and vulnerability.

AitĀ 3.2.2: Kauṇṭharavya's Teaching on "Connections" (sandhi)
As mentioned, the first two passages of AitĀ 3.2 are repetitive in content and structure and appear to present a variation on the same teaching given by different teachers. Where the first of these passages employs the term saṁhitā to describe the relationship of the devatā to the puruṣa at macrocosmic and personal levels (adhidaivata and adhyātman), the second passage replaces this with sandhi (written here saṁdhi), the term used to describe the euphonic combinations whose verbal production characterises the saṁhitāpāṭha mode of recitation. While the precise intention of the speaker cannot be determined, the shift from saṁhitā to sandhi draws attention to the way that parts of speech are depicted functioning as media connecting the person to the macrocosm. Note that the same connections (sandhi) account for both physiological vitality and religious attainment: when the identification of the self through speech is realized, one achieves outcomes that define the living of a complete or entire life (sarvam āyur). with respect to the self. For marrow is indeed this chief breath, this semen, and surely semen is not emitted without breath -if it were emitted without breath, it would putrefy, it would not yield. Those which we have called connections with respect to the deities are the joints with respect to the self. Of these three -of the bones, the marrows, and the joints -there are five hundred and forty [and] five [hundred and forty]; it becomes one thousand and eighty; one thousand and eighty indeed are the rays of the sun that change into the br̥ hatī verses and the day. The self that is composed of sight, composed of hearing, composed of the metres, composed of the mind, [and] composed of speech is this one measured with the syllables. He who knows thus this self as being measured with the syllables, composed of sight, composed of hearing, composed of the metres, composed of the mind, [and] composed of speech obtains intimacy, assimilation, and cohabitation with the syllables, becomes rich in offspring and livestock, he lives an entire life.
The numerical allocations in Kauṇṭharavya's teaching alert us to the fact that this passage contains a purposeful and focused development of the teaching given in the preceding. The 720 days and nights of the year that AitĀ 3.2.1 identifies with the 360 parts on each side of the human body are doubled here in AitĀ 3.2.2, firstly by positing the existence of 360 × 3 divisions of speech, and secondly by positing 540 parts on each side of the body. The resulting total of 1080 equates to the rays of the sun that this passage identifies with the br̥ hatī verses and each day. Again, where AitĀ 3.2.1 asserts that sight, hearing, mind, speech, the senses, and the body -namely, the entire self -are fixed together (samāhita) on the breath (prāṇa), AitĀ 3.2.2 shifts focus to the "chief" or "leading" breath (saṁpratiprāṇa) whose presence or absence leads respectively to fertility or decay. 27 Finally, in moving beyond the previous identification of the self with the days to instead posit that the self is like the syllables in number (akṣarasaṁmānaḥ…ātmā), Kauṇṭharavya's teaching draws the act of recitation -and, implicitly, speech itself -into the framework of connections placed around the identification of the self with the sun.
AitĀ 3.2.2 thus advances the key postulations of AitĀ 3.2.1 -that breath is the "beam" which supports the other devatā, 28 and that the self is like the days in number 29 -by expanding them to include the relationship between breath and speech, providing this with an anchoring in the success of the puruṣa's physiological functioning and asserting an identification between recitation and the sun. As we will see, this relationship of self-syllables-recitation-sun plays a critical role in the passages that follow.

AitĀ 3.2.3: Bādhva's Teaching on the "Four Persons")
These progressions reach a crux in AitĀ 3.2.3 despite this passage's impression of deviating from the preceding pattern of teachings. This is because unlike the initial two passages which take their orientation from the vertical hierarchy implicit in the identification between adhidaivata and adhyātman spheres, AitĀ 3.2.3 is characterized by a complex model that progresses vertically, with each successive level encompassing the one prior, but which ultimately inverts or reverses with the outermost and most expansive level being identified as the essence (rasa) of the innermost. This 'concentric' hierarchy is explained through the image of four persons (puruṣa) -the persons of the body (śarīra-), the metres (chandas-), the Veda (veda-), and the Great Person (mahāpuruṣa) -and it is the final identification of the Great Person with the essence of the śarīrapuruṣa that provides a coherent link to the restriction on recitation that concludes this passage.
Given the importance of AitĀ 3.2.3 to our ability to detect a logical sequence of contents in the SU's second adhyāya, it is worth considering in some detail.
The first stage of Bādhva's teaching introduces the four persons and addresses the śarīrapuruṣa and chandaḥpuruṣa specifically. While at first glance we might expect the movement from śarīra to chandas to implicate a step upwards in Bādhva's cosmological hierarchy, it is possible that the śarīrapuruṣa and the chandaḥpuruṣa are in fact presented as equivalent. I base this observation on the number of identifications that are made around each person: in contrast with the following two persons, which are each characterized by increases in their respective number of identifications, the śarīrapuruṣa and chandaḥpuruṣa each show two, equivalent statements -first, a bandhu disclosing identity, and second a statement of essence (rasa). On one hand these statements move in typically Upaniṣadic fashion from the least to most subtle aspects of each target. On the other hand, however, the structural symmetry seen here distinguishes the first two persons from the following. If my suggestion is accuratethat the śarīrapuruṣa and chandaḥpuruṣa are being treated as equivalent -then this introduction into Bādhva's teaching must build directly upon the identification of the self and the syllables with which the previous passage concluded, thus marking another progression in a coherent sequence.
There are four persons (puruṣa), according to Bādhva 30 -the person of the body, the person of the metres, the person of the Veda, and the Great Person. What we have called the person of the body is indeed this corporeal self (daihika ātmā); its essence is the incorporeal intelligent self (aśarīraḥ prajñātman). What we have called the person of the metres is indeed the traditional collection of syllables; its essence is [the phoneme] a. 31 30 catvāraḥ puruṣā iti bādhvaḥ śarīrapuruṣaś chandaḥpuruṣo vedapuruṣo mahāpuruṣo iti | śarīrapuruṣa iti yam avocāma sa ya evāyaṁ daihika ātmā tasya yo 'yam aśarīraḥ prajñātmā sa rasaḥ | chandaḥpuruṣa iti yam avocāmākṣarasamāmnāya eva tasyaitasyākāro rasaḥ | 31 This teaching also occurs at AitĀ 2.3.8, which contains an elaborate poetic and grammatical demonstration of the statement "brahman is a and the aham contained therein" (a iti brahma tatrāgatam aham iti). This passage features a significant ring-composition and utilises anagrams, metrical inversions, assonance, and etymological contrasts to build upon the statement cited.

The next stage of Bādhva's teaching introduces the person of the Veda (vedapuruṣa).
Here we observe an increase in the network of connections placed around the puruṣa -in addition to the two disclosures (viz., of identity and essence) seen previously, there is an additional pronouncement that builds out from the statement of essence. In this stage, the additional pronouncement regards the selection of a brahman priest or ritualist that is brahmiṣṭha, "replete in brahman": 32 What we have called the person of the Veda is that by which one knows the Vedas -the R̥ gveda, the Yajurveda, the Sāmaveda -its essence is the brahman priest. Therefore, one should choose a brahman priest that is replete in brahman, who would see a sacrifice's excess (ulbaṇam).
Again, it appears to be the focus on speech's role in the development of personhood that is the contextualizing framework for the progression seen here, moving attention from cosmological connections to the place of human agency. Working from the assumption that is through the Veda that the previous identifications between human embodiment and the syllables are revealed, this shift from the microcosm of the human person to the mesocosm of ritual practice might be interpreted as expanding inclusively around the śarīrapuruṣa and chandaḥpuruṣa.
However, the corollary introduction of the brahman priest is perhaps the more significant innovation. This is not simply because of the ritual agency that this figure implies -although this is both significant and relevant to the passage that follows (AitĀ 3.2.4) -but also because the figure of the brahman is open-ended. Unlike the other priestly roles involved in ritual performance, such as the hotr̥ , etc., which are predetermined in the sense of involving the delivery of undeviating contributions to a ritual, the brahman's identity is based on an understanding of sacred speech and their performance is presumably flexible and spontaneous. Through the introduction of this figure the SU posits a conceptual paradigm that emphasizes the importance of perceiving latent influences that threaten to destroy the balance of a ritual and thus the ability of a ritual to be productive but settles this as a quality of personhood.
Of particular interest in this regard is the employment of the term ulbaṇa, a relatively uncommon word that seems to refer to a conceptual category rather than to a precisely determined referent. Standard definitions of ulbaṇa stating it denotes the the amniotic membrane appear to derive from attestations in the Taittirīya Āraṇyaka (TĀ 1.10.7) which refer literally to calving in illustration of how the sun's rays are the caul produced from the pregnant night. 33 This meaning seems to be logically supported by the existence of a corresponding adjective, ulbaṇa, used to describe primary qualities of being massy, thick or large, and the secondary quality of being extraordinary. 34 However, it is the negated form anulbaṇa that may provide the key to understanding the employment of ulbaṇa in AitĀ 3.2.3. Occurring twice in R ̥ V, the 32 vedapuruṣa iti yam avocāma yena vedān veda r̥ gvedaṁ yajurvedaṁ sāmavedaṁ tasyaitasya brahmā rasaḥ | tasmād brahmāṇaṁ brahmiṣṭhaṁ kurvīta yo yajñasyolbaṇaṁ paśyet |. 33  adjective anulbaṇa is used to express the desire that an end product or procedure not have any excess or, perhaps, unwanted by-product. What seems clear in both of these occurrences is that the quality described as ulbaṇa is something that would disrupt the ideal state, function, or intended outcome of the phenomenon that is the focus of each verse: R ̥ V 10.53.6c: anulbaṇáṃ vayata jóguvām ápo 35 weave a work without knots for the ever-praising ones' 36 R ̥ V 8.25.9b: anulbaṇéna cákṣasā by means of eyesight without motes' 37 These attestations suggest that ulbaṇa may have been perceived as a flaw. In addition, Sāyaṇa's commentary on AitĀ 3.2.3 suggests that ulbaṇa is something inherent within or deeply related to (saṁbandhin) a phenomenon, process, or product -indeed, as a caul to a foetus -but that it is a morbid element with qualities of intensity or excessiveness. 38 On the basis of the examples cited, we might surmise that ulbaṇa (1) is generated or exists as a corollary of another action or phenomenon, is excess to want, and causes disruption; (2) that this disruption upsets the proper functioning or intended outcome of the related action or phenomenon; and (3) that the related action/phenomenon can obtain at any level of the cosmos -from childbirth to hymnproduction and sacrifice -but that its proper or ideal functioning is affective and relevant to ritual and the person alike.
I would argue that it is ulbaṇa rather than the more obvious shift from mesocosm to macrocosm that provides the core of the progression into the next and final stage of Bādhva's teaching. What this next stage signifies is not just the completion of the bandhus connecting the four persons, but a persuasive evolution from the factors that join (sam + dhā) within one's person and give rise to physiological identity, to a transpersonal phenomenon, the year, that sets the conditions both for life and for its failure to thrive. These conditions are encoded at the level of the individual person by the identification between prajñātman and sun -and we might recall that prajñātman was previously named as the essence of the śarīrapuruṣa in the first stage of Bādhva's teaching. The crux of the passage, Bādhva's revelation on saṁhitā, sees this relationship at all levels of creation: What we have called the Great Person is indeed the year that causes some beings to fall apart and others to aggregate; its essence is the sun. One should 35 Sanskrit text of the R ̥ V from Barend A. van Nooten & Gary B. Holland (1994). 36 Jamison & Brereton (2014, 1462. 37 Jamison & Brereton (2014, 1082. 38 AitĀBh on AitĀ 3.2.3: yajñasya saṁbandhinam ulbaṇaṁ tīvraguṇaṁ doṣaṁ… |.It is possible that these qualities of inherence and excessiveness are reinforced in AitĀ 2.3.8, in which the adjective ulbaṇiṣṇu is positioned in polar opposition to krūra ('harsh') and both are attributed as belonging to vāc; see Keith (1909, 117): yad vāca om iti yac ca. neti | yac cāsyāḥ krūraṁ yad u colbaṇiṣṇu | tad viyūya kavayo anvavindan). Keith's translation of ulbaṇiṣṇu as "immense" (1909,225) differs from Sāyaṇa, who suggests that ulbaṇiṣṇu refers to overtly conspicuous aspects of speech. See AitĀBh on AitĀ 2.3.8: ulbaṇiṣṇu atyulbaṇaṁ vācaḥ svarūpam ākrośādikaṁ…. know that this incorporeal intelligent self and the sun are one. Therefore, the sun [appears] equally to each and every person. This was said by a r̥ ṣi: "The brilliant face of the gods has arisen, the eye of Mitra, Varuṇa, and Agni; he has filled heaven, earth, and the space between: the Sun is the life-breath (ātmā) of both the moving and the still." 39 Thus, Bādhva has said "I consider this saṁhitā as being put together according to the correct order." For, indeed, the Bahvr̥ cas examine this in the great hymn, the Adhvaryus in the fire, the Chandogas in the Mahāvrata ritual; [they examine] this in this [world] -this in heaven, this in the wind, this in space, this in the waters, this in the plants, this in the trees, this in the moon, this in the stars, this in all beings -this alone they call brahman. The self that is composed of sight, composed of hearing, composed of the metres, composed of the mind, [and] composed of speech is this one measured with the year. 40 Bādhva asserts that saṁhitā is not just sandhi or a mode of recitation, but the truthful identifications, the state of being 'put together', that recitation reveals. This state of saṁhitā follows the structural order contained in the identification of self and sunan assertion that has multiple implications for our understanding of the SU. Firstly, it suggests that saṁhitā is conceived of as a natural state of affairs. To the extent that this relies largely upon the sun as the essence of the year, it implicates both the potential for life and, by extension, the possibility of collapse as natural conditions at play in the individual person and the world. Secondly, it implicates saṁhitā in the transpersonal identity that is an outcome of true knowledge but brings the puruṣa into focus as an object and means of self-realization. This contrasts significantly with widespread popular notions of Upaniṣadic teaching that emphasize knowledge of the ātman and depict the puruṣa as having little relevance to self-realization. Yet, as we shall see, the implication of the puruṣa in AitĀ 3.2.3 establishes a coherent religious context for the following passage, AitĀ 3.2.4, which is highly illustrative of the kind of content that has been seen as diverging from the Upaniṣads' metaphysical focus.
Thirdly, while Bādhva's declaration extends the meaning and applicability of saṁhitā well beyond simple reference to the recitation of the R ̥ V, it does so without breaking the underlying relationship between speech and the puruṣa seen depicted elsewhere in the SU. Because of this speech, saṁhitā, personhood, and knowledge share conceptual terrain and structural similarities, but in what follows the SU sug-39 R ̥ V 1.115.1 in translation by Jamison & Brereton (2014, pp. 266-267 gests that it is the knowledgeable relationship to speech that determines the personal outcomes of recitative practice.

AitĀ 3.2.4: the reversal of saṁhitā
Rather than addressing the benefits to be gained from knowledge of saṁhitā, AitĀ 3.2.4 focuses on saṁhitā's undoing and the puruṣa's collapse -unsurprisingly, its contents are far removed in tone from the preceding passages. These features do not fit easily with widespread conceptions of either saṁhitā or Upaniṣadic metaphysics and, as stated previously, this passage has certainly contributed to impressions of the SU's miscellaneous nature. Nonetheless, this passage arguably contains some of the most valuable resources for our understanding of the SU. This is firstly because its focus on what saṁhitā is not highlights the scope of what saṁhitā is -this assists us in developing an outline of saṁhitā in relation to the SU's depictions of personhood and the impacts of recitation. Secondly, because recitation is now depicted as having a destructive result when it is practiced improperly, the vulnerability of the puruṣa is brought to the fore. And this in turn alerts us to a concern shared by all of the passages examined here, namely, with the relevance of saṁhitā to the development and preservation of personal identity.
In contrast with what precedes it, the teaching that occupies AitĀ 3.2.4 commences in the final line of AitĀ 3.2.3, replacing the expected statement that one "who knows thus this self as being measured with the syllables, composed of sight, composed of hearing, composed of the metres, composed of the mind, [and] composed of speech obtains intimacy, assimilation, and cohabitation with the syllables, becomes rich in offspring and livestock, he lives an entire life." 41 More than simply marking an inconsistency in the text's sequencing, as Keith understood it (1909, p. 251, n. 7), the placement of this shift emphatically redirects the focus of the teaching from 'one who knows thus' to one who does not. This is initially formulated as a restriction on recitation according to which the repercussions of improper recitative practice result in the loss of the positive relationship between the puruṣa and speech that was the basis for the achievement of saṁhitā in the preceding teachings.
The self that is composed of sight, composed of hearing, composed of the metres, composed of the mind, [and] composed of speech is this one measured with the year. 42 For this one, the one who thus recites for another this self as 41 Repeated with minor variation as the concluding statement of AitĀ 3.1.1 and 3.1.2: sa eṣa saṁvatsarasaṁmānaś cakṣurmayaḥ śrotramayaś chandomayo manomayo vāṅmaya ātmā | sa ya evam etam akṣarasaṁmānaṁ cakṣurmayaṁ śrotramayaṁ chandomayaṁ manomayaṁ vāṅmayam ātmānaṁ vedākṣarāṇāṁ sāyujyaṁ sarūpatāṁ salokatām aśnute putrī paśumān bhavati sarvam āyur eti || The first line of the formula occurs as part of the conclusion of AitĀ 3.1.3, but the second line (sa ya evam…sarvam āyur eti) is replaced by the restriction on recitation that leads into AitĀ 3. being measured with the year, composed of sight, composed of hearing, composed of the metres, composed of the mind [and] composed of speech [AitĀ 3.2.3], the Vedas are milked dry, there is no share in the oral tradition, he does not know the path of the rightly performed. This was said by a r̥ ṣi: "Who has abandoned the companion joined (to him) in knowledge, for him there is no share in Speech at all. When he hears her, he hears in vain, for he does not know the path of the rightly performed (ritual)." 43 This is said -that he has no share in the oral tradition, he does not know the path of the rightly performed. Therefore a wise one would not pile the fire for another, nor praise with the Mahāvrata for another, nor recite this day for another. 44 Only for a father or a teacher may one recite freely, for that is done for oneself…. [AitĀ 3.2.4] The next stage of this passage describes the loss of the relationship between the sun and the self. It seems likely that this is anticipated by Bādhva's declaration that the year, whose essence is the sun, is that which causes some beings to fall apart and others to aggregate, now framing it within an extended context of verbal and ritual conduct introduced by the restriction on recitation. However, where Bādhva's teaching employs the verb pra + dhvaṁs to express the state of falling apart or perishing, the current passage utilises the verb vi + hā. This could well be inconsequential, yet it is feasible that this is an intentional change of vocabulary that nuances the impact of the description contained in this passage and signifies another progression in a conceptually related sequence. This is because while vi + hā is widely attested in Sanskrit literature in the sense of the breaths' departure from the body, thus indicating not only the death of the corporeal self but also the means by which one steps forth into postmortem immortality, in this instance it describes a loss of identity rendering such a departure from the body impossible.
This raises some significant considerations for the way the next stage of this passage is interpreted. The fact that the ultimate identity of the self and the sun is depicted as something that can be lost suggests that the SU conceives of this relationship as violable. This should prompt us to ask how such violability might be encoded within the SU's metaphysical outlook and whether it establishes a sufficient basis for rites aimed at self-protection to be seen as a coherent component in the Upaniṣads' practical emphasis on the achievement of liberating insights. Following this, it is evident from the next stage of the passage that the loss of this ultimate relationship disrupts the functioning of the puruṣa, that is, of the collocation of devatā whose combined parasmā agniṁ cinuyān na parasmai mahāvratena stuvīta na parasmā etad ahaḥ śaṁset | kāmaṁ pitre vācāryāya vā śaṁsed ātmana evāsya tat kr̥ taṁ bhavati | … || AitĀ 3.1.4 43 R ̥ V 10.71.6 in translation by Jamison & Brereton (2014, p. 1497); see also their discussion of R ̥ V 10.71 (pp. 1496-97). 44 The parallel statement in the Śāṅkhyāyana Āraṇyaka (ŚāṅkhĀ) contains interesting variations on this restriction. See ŚāṅkhĀ 8.6: nāsyānūkte vāco bhāgo astīty eva tad āha tan na parasmā etad ahaḥ śaṁsen nāgniṁ cinuyān na mahāvratena stuyān ned ātmano 'pihīyā iti ("His speech has no share in the oral tradition -so it is said. Thus, he should not recite this day for another, nor pile the fire, nor praise with the Mahāvrata lest he becomes detached from his self.") contribution establishes a matrix for developments of personhood (i.e., for knowing that the self composed of sight, hearing, metres, mind, and speech is measured with the syllables, etc.). Presumably, this combined contribution is also prerequisite to achieving the goals that the SU repeatedly identifies (i.e., intimacy, assimilation, and cohabitation with the syllables, richness in offspring and livestock, and living an entire life). Thus, we see that improper verbal conduct compromises saṁhitā, whether it is taken as a natural phenomenon or an achieved state of personhood, and that this bears strongly upon an inherent relationship between saṁhitā and speech. The affective impact of speech on the puruṣa becomes increasingly pronounced in the next stage of the passage: We have said that this incorporeal intelligent self and that sun are one. 45 When they are lost (vihīyete), the sun looks like the moon, its rays do not manifest, the sky turns red as madder, [his] rectum prolapses, 46 his head emits the smell of a raven's nest [and] he should know that his self is gone, he will not live for long. He should do what he considers needs doing; [following that] he should silently recite the seven verses beginning "What [peril] is near and what is far," the single verse "Of the age-old semen," the six verses "Where, o self-purifying one, the formulator," [and] the single verse "We, up from the darkness to the higher [light]." 47 That the malfunction described here involves a disintegration of identity is suggested firstly by the odours and winds that break through the subject's bodily boundaries. As mentioned above, it seems likely that this indicates the destabilization or departure of the devatā that are otherwise established (pratiṣṭha) within their respective corporeal abodes (āyatana) and whose integrated departure from the body at death, via prāṇa, is a recurring theme in both Āraṇyakas of the R ̥ V. 48 This is suggested again by the selection of verses prescribed, which seem intended to mitigate against the loss of a puruṣa that is no longer functional, and hence has ceased to be capable of achieving personally transformative aims. While not returning the subject to the state of integration and abundance in the world that was described in preceding passages 45 sa yaś cāyam aśarīraḥ prajñātmā yaś cāsāv āditya ekam etad ity avocāma | tau yatra vihīyete candramā ivādityo dr̥ śyate na raśmayaḥ prādur bhavanti lohinī dyaur bhavati yathā mañjiṣṭhā vyasthaḥ pāyuḥ kākakulāyagandhikam asya śiro vāyati saṁpareto 'syātmā na ciram iva jīviṣyatīti vidyāt | sa yat karaṇīyaṁ manyate tat kurvīta yad anti yac ca dūraka iti sapta japed āditpratnasya retasa ity ekā yatra brahma pavamāneti ṣaḷ ud vayaṁ tamasas parity ekā | 46 Possibly, the wind is not retained in the body. This seems to refer to the malfunctioning of prāṇa, which, as we have seen, is elsewhere described as supporting all other devatā in the puruṣa and thus ensuring its correct functioning. 47 These pratīkas reference, in order of citation, R ̥ V 9. 67-21-27, 8.6.30, 9.113.6-11, and 1.50.10. Interestingly, all of the verses prescribed for silent recitation are characterized by the peculiarity of their respective embedding within entire hymns of the R ̥ V, being either self-contained units, enigmatic transitions or intermediate points between more clearly defined groups of material, or climactic omphalos verses. With respect to the designations of the hymns themselves, the first and third are dedicated to Soma Pavamāna, the second to Indra, and the fourth to Sūrya, thus resonating predictably with the themes of the Mahāvrata. The translation of the pratīkas is adapted from Jamison & Brereton (2014). 48 For example, AitĀ 2.1. 4-5, 2.3.3, 2.6; ŚāṅkhĀ 4.12-14. with their focus on the meaning of sam + dhā, these recitations ostensibly counteract the devastating repercussions of vi + hā by ensuring entry into immortality after death. Indeed, each of the four recitations prescribed can be seen to correspond to one stage in a consecutive sequence of transactions between the puruṣa and vāc: the first stage, instigated by R ̥ V 9.67.21-27, involves the purification of the person through the coded characteristics of a variety of gods; the second, instigated by R ̥ V 8.6.30, re-establishes the power of one's own speech-acts as substances fit for offering; the third, instigated by R ̥ V 9.113.6-11, harnesses this re-invigorated speech to instil Soma (and hence the fluid of immortality, amr̥ ta) with the ability to transport the speaker from the ritual sphere to heaven; and the fourth, instigated by R ̥ V 1.50.10, aims to reconnect the subject with the sun, the "highest light" and the truth of their identity. 49 The passage then turns to a lengthy enumeration of sensory hallucinations (pratyakṣadarśana) and dreams (svapna), both traditional portents of affliction. These symptoms again function as indicators of the puruṣa's malfunction -the body's winds, traveling from the inside out, are now followed by misperceptions of external phenomena whose stimuli travel from the outside in. This can be seen as a reversal of the physical conditions that previously accompanied the realization of saṁhitā in the functioning of the puruṣa and its relational identity with the external world.
Next, if the sun looks split, like the nave of a chariot-wheel, or he sees his shadow is split, he should know that it indeed is so. 50,51 Next, if he sees himself in a mirror or in water with a crooked head or without a head, or if his pupils appear inverted or crooked, he should know that it indeed is so. Next, he should cover his eyes and look: circular lines of light that seem to fall together are visible. When he does not see them, he should know that it indeed is so. Next, he should cover his ears and listen: there is a sound such as of a blazing fire or of a chariot. When he does not hear it, he should know that it indeed is so. Next, when fire appears blue as a peacock's neck, or of he sees lightning in a cloudless sky, or doesn't see lightning in a cloudy sky, or sees what looks like glittering 49 Jamison & Brereton (2014, p. 163): "Looking up from the darkness to the higher light, we have come up to the Sun, god among the gods, the highest light." 50 According to the ŚāṅkhĀ parallel, what should be known is that one's self is gone and they do not have long to live; see ŚāṅkhĀ 8.7: saṁpareto 'syātmā na citam iva jīviṣyatīti vidyāt (cited in Keith 1909, p. 313 specks in a great cloud, he should know that it indeed is so. Next, when he sees the earth as if burning, he should know that it indeed is so. These are [the signs] perceived with the senses. Now, the dreams. He sees a black man with black teeth; that man kills him. A boar kills him. A monkey leaps at him. A fast wind blows him along. He chews on gold and spits it out. He eats honey. He devours stalks. He carries a single lotus. He travels with a team of asses and boars. Wearing a garland of red hibiscuses, he drives a black cow with a black calf south. If he should see any of these, then, having kept a fast, he should heat the milk for the offering of a sthālīpāka and offer it with the Rātrī hymn -an offering to a verse -and feed the Brahmans with other food and eat the oblation himself. He should know that the person within all beings who is not heard, not reached, not thought, not seen, not discerned, [and] not determined, [but is] the hearer, the thinker, the seer, the determiner, the sounder, the discerner, [and] the knower is his self. A closer examination of the sthālīpāka offering prescribed in this passage allows us to see how it responds to the SU's earlier depictions of saṁhitā and the dynamics it involves. This relates particularly to the role of external influences which are integrated within the puruṣa, giving rise to its functioning and development. Although the utilization of sūktas (or parts thereof) for the purpose of warding off nightmares and neutralizing their negative influences was established practice by the time of the Gr̥ hyasūtras, this passage appears to be unique in its deployment of the Rātrī hymn (R ̥ V 10.127) together with the production of a sthālīpāka, the "most characteristic of domestic offerings" (Gonda, 1980, p. 190). In contrast with Jan Gonda's interpretation of this passage, 52 however, I note that the hymn to Rātrī is neither concerned with the fearsome darkness described elsewhere in the R ̥ V nor asks protection from ill-dreams and omens, but celebrates an immortal goddess smeared over with stars who repels the dark by sending forth her sister, Dawn. As such, while this hymn evokes the night as a space of comfort and rest, its poetic function anticipates the return of the sun, 53 and we might speculate that the restoration or reintegration of the subject's true identity is what this particular combination of recitation and sthālīpāka is intended to effect. The intertextual documentation of sthālīpāka offerings points consistently to an understanding that the combined involvement of the deity to whom the offering is dedicated and the family-or community-members who typically share consumption of the oblation generates an extended relationship between the one who makes the offering and the event precipitating it. The sthālīpāka thus functions as a transactional strategy: 54 the simultaneous offering of hymn and oblation is considered to imbue the oblation with the objectives set forth in the hymn, and this newly constructed identification is incorporated into the persons who consume the oblation (Marriott 1976, pp. 114, 136). 55 Of all the passages examined here, AitĀ 3.2.4 draws particular attention to the influence of traditional notions about personhood within the SU's teachings. This is because the loss of identity and state of collapse described in this passage draw their coherence from the puruṣa's formation with the devatā and the volitional dynamics the latter display. The ritual practices prescribed by this passage also draw coherence from this same context. According to this framework, inappropriate ritual conduct does not simply fail to yield desired outcomes but compromises the network of relationships that connect the puruṣa to the external world and fractures the connection of the self with its ultimate identity. We might conclude that these factors inform a context in which the personal states associated with saṁhitā -namely, being put together and thriving -are urgent and practical desires addressed by means of recitation. Following this, it is apparent that the symptoms and rituals described in AitĀ 3.2.4 progress the discussion of saṁhitā contained in the preceding passages, albeit through the lens of saṁhitā's reversal, in a way that clarifies the factors and dynamics previously raised by the text.

Conclusions
At the beginning of this examination, I raised the question of how previous interpretations of saṁhitā may have influenced perceptions of the SU as being miscellaneous in character and lacking internal consistency. A key factor in this consideration was the need to closely examine the dynamics and concerns placed around saṁhitā in the SU, which I suggested would elucidate saṁhitā's scope and thus allow us to better identify underlying points of convergence in one of the SU's most diverse and therefore most challenging textual sequences. I also contended that this would require an inclusive approach, placing priority with the form of the text as it has traditionally been transmitted rather than excluding content that at first glance appeared to be unrelated to the SU's stated focus on saṁhitā.
This brings us to a pair of related conclusions. The first of these is that saṁhitā describes a quality of creation or manifestation, and it implies a structural order of shared identifications that connect the person with the external world. It is also descriptive of an achieved state of personhood, in which respect it links the personal realization of saṁhitā within oneself with the attainment of lived outcomes that in some sense promote the state of vitality that the SU associates saṁhitā with. We might thus say that saṁhitā is not so much a quality of recitation as a phenomenon that recitation instantiates and reveals. In order to see this it has been necessary to consider recitation in the context of its verbal, physiological production rather than in the abstract. Put differently, the SU carries a reminder that while it is possible for us to think about recitation without considering its physical production, recitation as a ritual practice aimed at producing personal outcomes is not easily divorced from notions of how a person functions. Indeed, in the SU it is the latter that provides a conceptual framework accommodating the anticipated impact and relevance of the former.
Following on from this, a second conclusion points to the correlation between the SU's depictions of saṁhitā and the internal relationship between the SU's contents. We have seen that the SU develops a picture of saṁhitā that addresses it in an extended context of considerations involved in its realization as a personal achievement. This includes warnings about the repercussions of improper recitative conduct, which are likewise positioned within the framework of the puruṣa's functioning. While the SU's contents are certainly diverse, they share a common core which accounts for its shifting focus and suggests that the concern with self-preservation is warranted by the SU's metaphysical outlook. From these considerations, the SU emerges as an internally coherent Upaniṣad containing a related sequence of teachings. These findings establish a basis for the more detailed study of the SU in its entirety and suggest an approach to the wider examination of the R ̥ gvedic Āraṇyakas and similar Vedic texts that demonstrate even higher levels of internal diversity.

Authors' Contributions
This article is the work of a single author, Stephanie Majcher. No other authors were involved in material preparation, data collection and analysis, and the production of manuscript drafts or the final manuscript.
Funding The author has no relevant financial or non-financial interests to disclose.

Open Access funding enabled and organized by CAUL and its Member Institutions
Availability of Data and Material Not applicable.
Code Availability Not applicable.

Conflicts of Interest/Competing Interests
The author has no conflicts of interest to declare that are relevant to the content of this article.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/ licenses/by/4.0/.