Being Is Relating: Continuity-in-Change in the Sambandhasiddhi of Utpaladeva

Abstract

Relation-theories—theories on the metaphysical status of relations—have for some time stood at the center of disputes between realism and idealism. To such disputes, this paper contributes insights from an understudied premodern source, the Sambandhasiddhi (Proof of Relation). Its author Utpaladeva (c. 925–975 C.E.) is the Śaiva philosopher of India best known as an innovator in the Pratyabhijñā (Doctrine of Recognition) school of Kashmiri Śaivism. This lesser-known late text shows Utpaladeva deploying an even more explicitly Bhartṛharian grammatical view of reality than he had previously. He argues against his chief rival and predecessor, the Buddhist epistemologist, Dharmakīrti (c. 6th or 7th C.E.), while modifying the latter’s epistemic idealism to an objective idealism. This text differs from Utpaladeva’s prior works in its sustained attack on Dharmakīrti’s nominalism and citation of the Buddhist’s own writings. The Sambandhasiddhi accordingly offers an interesting glimpse at a sustained treatment on relations, a topic that is important to Utpaladeva’s prior arguments, but that he considered perhaps not sufficiently developed, so as to warrant a separate treatment. A few brief comments are also offered on how Utpaladeva’s relation-theory might fit alongside Russell’s disputes with Bradley over relations, and Utpaladeva’s affinity with Peircean semiosis.

1. Introduction

Relation-theories, those concerning the metaphysical status of a relation between one object of perception and another, have long been tied to the tension between realist and idealist worldviews. Those skeptical about real relations tend to incline towards a sort of idealist subjectivism, perhaps motivated rightly enough as an antidote to a mechanistic naïve realism. But then still further, as a guard against the pernicious, solipsistic extreme of subjectivism, it can likewise be argued that relations, though not existents (something with independent status), are yet reals (something both persistent and reliable). That is, though a relation is not possessed of the same order of reality as the relata it connects, it has a sort of intersubjective existence. Though not something that exists (i.e., stands apart), it is real because it may reliably, and with a certain degree of precision, be agreed upon. A relation is then a consensus truth, capable of producing determinate, reliable results. Relation-theories are central to this kind of tension between what is empirically observable and what is subjectively determined.

Despite the all but disappearance from Western philosophical discourse of serious champions of idealisms, relation-theories bear upon the tension between subjectivist and objectivist ways of experiencing the world. This tension remains in place even up to the present day. It becomes a central dilemma in, for example, attempts to interface non-reductively the subjective social sciences and objective physical sciences in transdisciplinary cooperation. As Bertrand Russell comments on empiricism:

The observer when he seems to himself to be observing a stone, is really, if physics is to be believed, observing the effects of the stone upon himself. Thus science seems to be at war with itself: when it most means to be objective, it finds itself plunged into subjectivity against its will. Naïve realism leads to physics, and physics, if true, shows that naïve realism is false. Therefore naïve realism, if true, is false; therefore it is false.

(Russell [1940] 1965, p. 13)

Not just the physical sciences, but the social sciences encounter this problem, only in reverse. Understanding the nature of one’s own subjective awareness is tricky, as it cannot but take subjective awareness as both the object of investigation and the instrumental means to do so. The well-known hard problem of consciousness (Chalmers 1995) has received interesting treatment recently from Mark Solms (Solms 2021), who argues for consciousness as having its root in affective, emotional states.1

Transdisciplinary or interdisciplinary methods of interfacing social and physical sciences have thus stood in need of ways to share vocabularies non-reductively, a process of which theories of relation may play an important part. The interfacing of transdisciplinary or interdisciplinary frameworks is a process admirably carried forward recently by Soren Brier’s Cybersemiotics (Brier 2008), which is part of a broad effort to develop the many promising implications of posthumously organized and published works of C.S. Peirce along with Hoffmeyer’s Biosemiotics (Hoffmeyer 2008) and others. While Peirce lays the groundwork for rich discussions on relations and their role in what is for him an intimate connection between truth and widespread consensus, there has also arisen recently some reason for a fresh look at 20th century skeptical attacks on relations and the responses to them from within the analytic tradition.

Russell’s chief rival in the above sort of tension between the subjective and objective, F.H. Bradley, attacks real relations with his famous regress argument: Relations are either separate from their relata or one with them (Bradley [1893] 1962). That is, they may be separate entities, or they may be properties. If truly separate, relations would need sub-relations to connect the relata to their relations, as well as sub-sub-relations to relate these sub-relations, etc., in a vicious regress. Another version of the same problem arises if relations are one with their relata, as properties. There, sub-relations would seem to be required to hold the property as distinct from its ground, and so on. Russell’s direct response to this regress, it has recently been shown by Candlish, is not entirely adequate. Russell fares somewhat better when not facing Bradley head-on in making a defense of relation, but instead adopting a linguistic approach to empiricism. Peirce’s thoroughly semiotic view of reality fares better still. Peirce’s semiosis amounts to a convincing defense of relation against metaphysical idealism or nominalism. A regress argument like Bradley’s, and a defense similar to that of Peirce, however, have much earlier and fascinating precedents in South Asia, in the tension between Dharmakīrti (c. 6th or 7th C. C.E.) and Utpaladeva (c. 925–975 C.E.). The following treatment of Utpaladeva, then, incorporates Western philosophers not because any Western model is necessary. On the contrary, my basic assumption is that Utpaladeva brings things to the table undreamt of in Western philosophy, and that his inclusion in a contemporary philosophical discussion is of benefit.

The Nominalist Repudiation of Relation

The purpose of the present work is to consider a much earlier version of the subjectivist vs. objectivist debate in the late work of the Śaiva philosopher Utpaladeva, and the roots of its implications in organizing knowledge and informing ethical dispositions across disciplines. Utpaladeva’s chief rival, the Buddhist epistemologist Dharmakīrti, puts forward a regress argument that, though more laconically communicated, is essentially identical to the one Bradley proposes. Utpaladeva’s is an early example of how to avoid engaging an opponent’s skepticism head-on, and he fares much the better for it. Rather than push back against Dharmakīrti’s regress argument itself, Utpaladeva takes aim instead at the roots of his opponent’s own worldview: a skeptical reductionist take on language.

As is amply demonstrated by Torella ([1994] 2013), nominalistic relation-theories are central to the Dharmakīrtian, Buddhist epistemologist deconstruction of persons as enduring selves. It is by dissolving any substantial, mind-independent inner relations within the person that the idea of the person as an enduring self is dissolved. The attack on the ātman of the person is just one important special case of the Buddhist attack on permanent and independent selves of perceptual objects generally. Dharmakīrti’s analysis, in its most subtle level, caches out in a state of neither realism nor metaphysical idealism. Rather, it is a sort of epistemic idealism (Dunne 2004); it neither affirms the essential self of the philosopher, nor indeed the essential identity of any objective self. All that remains, in this view, is a realm of pure appearance, or manifestation. But securing this kind of pure epistemology (a realm not of real things but of pure manifestation) comes at an important ethical cost. Ratié has elaborated upon Śaiva critiques of the weakness of Dharamakīrti’s system as “incapable of accounting for the mundane experience of otherness.” (Ratié 2007, p. 317). Not a universal statement of Buddhist selflessness, but rather a kind of solipsism, the Dharmakīrtian vision is one in which the existence of other sentient beings is less secure than the philosopher’s own existence. The philosopher’s subjective personhood is a secure epistemic ground. Sentient beings are only ephemeral objects of the philosopher’s perception. It is not simply that other sentient beings, like the philosopher, have no absolute or independent existence. Rather, other sentient beings only appear to exist. The Buddhist balance of compassion and wisdom is undermined, by its own standard. If the subjective existence of the philosopher is retained, with other sentient beings only mere appearances, why teach Dharma to begin with? Who would be there to hear it?

Such relation-theory, then, might be described as not only objectionable to its opponents, to those defending the person as enduring self. It also seems to fail by its own Buddhist standard of upholding both compassion and wisdom, since it cannot adequately account for other sentient beings.2 A direct rejoinder to the nominalist relation-theory comes from the Śaiva philosopher Utpaladeva, with his development of a type of objective idealism, which affirms the enduring self of the philosopher and sentient beings in general as both partaking of an omnipresent consciousness, a transcendental subjectivity, Śiva (Lawrence 1996, 1999, 2018, 2019). The Dharmakīrtian discontinuous world of pure manifestations is corrected and elaborated upon, becoming a world not of a dualistic split between manifestation and language, but a thoroughly semiotic continuum of continuity-in-change, and reality-in-manifestation.

Not surprisingly, then, Utpaladeva’s theory of relation is central to his world-affirming body of philosophy, just as Dharmakīrti’s skeptical refutation of relations is central to his world of mere manifestations. Utpaladeva’s best known work is his Īśvarapratyabhijñākārikās and coeval vṛtti (hereafter ĪPK, ĪPKv, respectively), where a relation argument is offered in the bhedābheda (unity-in-diversity) chapter (ĪPK 2.2, see Torella [1994] 2013). In his late text the Proof of Relation (Sambandhasiddhi, hereafter SS), Utpaladeva affectionately refers back to that earlier corpus simply as his Pratyabhijñā. In the SS, Utpaladeva embarks on an even more confident, almost peremptory refutation of his Buddhist opponent’s theory, a technique which, as I argue elsewhere, more closely resembles what in contemporary philosophy is called a “retorsive transcendental argument” than anything else Utpaladeva writes (MacCracken 2021). Such argument turns the Dharmakīrtian worldview in on itself. In doing so, Utpaladeva’s argument exposes the transcendental necessity of the relation the Dharmakīrtians purport to refute. Beyond revealing the inherent contradiction in his opponents’ argument, Utpaladeva argues emphatically for a unitary nature of relatedness in the absolute sense, and positively affirms the power to encompass diverse facets of experience within a single vimarśa (reflective awareness) in the relative sense. In what follows, I first outline my findings concerning how continuity-in-change represented in the SS constitutes an embrace of a neo-Bhartṛharian semiosis that is still more detailed than that of the ĪPK and ĪPKv (Section 2). I then offer some suggestive comments on how Utpaladeva’s theory of relation in the SS anticipates contemporary concerns, and how its translation and circulation might contribute to present discussions (Section 3).

2. Distinctive Features of the SS and Neo-Bhartṛharian Semiosis

Critically responding to the Sambandhaparīkṣā (hereafter SP, translated in Jha 1990) of Dharmakīrti while both recovering and building upon the Sambandhasamuddeśa (hereafter SSam, translated in Houben 1995) of Bhartṛhari, the SS of Utpaladeva refutes the nominalist inclinations of Dharmakīrti, and in the process affirms an objective idealism against Dharmakīrti’s epistemic idealism. As I have touched on previously, the SP begins from the realm of experience Dharmakīrti privileges: the epistemic world of sensation, using the Buddhist frame-by-frame analysis of kṣaṇikatvā (momentariness). Its result is a condemnation of the realm of inference and language, which Dharmakīrti seeks to subordinate to (purportedly non-linguistic) pure sensation.

Utpaladeva’s entire methodology is in the reverse. He means to show that relation is nothing less than a subtle capacity of an omnipresent universal consciousness, Śiva. Moreover, he means to show how the Dharmakīrtian epistemology already tacitly admits as much, or ought to do so, because their own philosophy wants dualistically to divide the world into true sensation and untrue inference. Yet, they neglect in their relation-theory adequately to account for the practical utility and reliability of relations in the manifest world, as opposed to the total lack thereof that would apply to total illusions. Even in the Dharmakīrtian’s own error-theory, Utpaladeva argues, there is apparent inconsistency with the dualist world their relation-theory would suggest. Their world of pure manifestation admits of phenomena such as what they call avayavins (part-possessors), which may be error from an absolute point of view, though not at all in the worldly sense. Relations are, Utpaladeva argues, much the same (MacCracken 2021).

The latest in history of the trio of sambandha-texts, the SSam, SP, and SS, the latter also represents Utpaladeva’s perspective later in the development of his thought, and indeed it may be the latest of his extant works. As I already mentioned above, it references both his ĪPK and ĪPKv, affectionately dubbing them together The Pratyabhijñā, and it also mentions by name his ṭīkā, or vivṛti (the ĪPKvv). In some of its vocabulary and style of argument it also bears resemblances to Utpaladeva’s commentary on the Śivadṛṣṭi (hereafter ŚD) of Somānanda (translated in Nemec (2011, 2021)), but does not explicitly mention the latter, and so may either predate or antedate that text. Of the Sidditrayī of Utpaladeva to which it belongs, Shastri is almost certainly correct that the traditional manuscript ordering (first the Ajaḍapramātṛsiddhi (Lawrence 2009), then Īśvarasiddhi (see Ratié 2015), then SS), also reflects their order of composition by the author (Shastri 1921, pp. 7–9). While I in no way intend to suggest a more refined Utpaladeva may be found in the SS, a study of it nonetheless helps flesh out an important picture of a latter phase of his thought. Owing to its siddhi genre (i.e., a short proof), the SS could serve as an introductory digest to his thoughts on the topic of relation. The text has a degree of depth, however, that also makes it an advanced study in Utpaladeva’s thoughts on this specific topic.

2.1. The Nature of Utpaladeva’s Neo-Bhartṛharianism

By way of background, it should be noted that Bhartṛhari’s conceptual tools exert a strong influence on the Pratyabhijñā doctrine beginning with Utpaladeva’s articulation of it. As Torella puts it, “The omnipervasiveness of language is an epistemological version of the omnipervasiveness of Śiva, and at the same time calls for the integration into the spiritually dynamic Śaiva universe.” (Torella 2008, p. 521) At the same time, it should not be forgotten that Dharmakīrti himself, in contrast with his predecessor Dignāga, displays what I would describe as an eisegetical (creative exegetical) relationship to Bhartṛhari. As Herzberger concludes her volume on the subject:

Dignāga interpreted Bhartṛhari’s conclusions as a challenge to his Buddhist commitments. He attempted to show that there is a fragment of language which can be built “in dependence upon” the knowledge received from the senses. Dharmakīrti, on the other hand, conceded Bhartṛhari’s claim that all our knowledge about external reality rests in language which is prior to experience. While Dignāga freed himself from Bhartṛhari’s conclusions by means of logic and his techniques of apoha, Dharmakīrti purchased at least part of his freedom by re-labeling essential aspects of Bhartṛhari’s doctrine of essential objects (what is nitya in Bhartṛhari becomes a product of an anādivāsanā for Dharmakīrti).

(Herzberger 1986, p. 241)

Dharmakīrti’s concession to which Herzberger refers concerns, in part, Dignāga’s antagonism with Bhartṛhari as contrasted with “Bhartṛhari who located the relation in ideal sentential objects, and Dharmakīrti who followed Bhartṛhari’s lead.” (p. 234). In the SS, Utpaladeva makes much of ideal sentential objects, as his task is to outline a list of diverse grammatical examples of relations in sentences, before then turning to consider whether the relations described have any real status. At stake also is what has been identified since Stcherbatsky’s Buddhist Logic as the Kantian a priori (Stcherbatsky [1964] 2004). As Herzberger continues:

The a priori is part of Bhartṛhari’s doctrine, the source of his idea that the Brahman is at the root of our understanding. Dharmakīrti, I have argued, adopted the a priori from Bhartṛhari, but consigned it to the realm of illusion, an anādivāsanā.

(Herzberger 1986, p. 234)

The Kantian interpretation of Dharmakīrti is a topic I take up in detail elsewhere (MacCracken 2021). It should hopefully suffice to say that the a priori here is Westernized description of various ways of drawing inferences that rest upon reason without being contingent on any empirical variables. Dharmakīrti famously privileges such reason while also, as Herzberger points out, insisting upon its strictly unreal status. Utpaladeva in his Pratyabhijñā first incorporates and presupposes much of Dharmakīrti’s doctrine before exposing its internal contradictions, at Torella demonstrates (Torella [1994] 2013).

In fact, given such strong inclusion and modification of the Dharmakīrtian, it may be tempting to propose, as a senior colleague of mine once quipped, that the Kashmiri Śaiva authors would be Dharmakīrtians if not for Bhartṛhari. However, the complete picture is of course more nuanced. Dharmakīrti is himself a reader of Bhartṛhari, albeit, in his nominalism and anti-eternalism, one who reads against-the-grain. Accordingly, the distinction between Dharmakīrtian and Utpaladevean philosophy is an important though fine-grained distinction, between contrasting idealisms as well as between contrasting doctrines of all being as constituted by appearance. It is chiefly the act of pursuing of these threads: idealism and a doctrine of appearance, that makes Utpaladeva’s philosophy what I would describe as an innovation as opposed to a revival (i.e., a neo-Bhartṛharianism as opposed to paleo-Bhartṛharianism). Indeed, such Kashmiri Śaiva innovations exerted an influence of their own back upon the grammarians. For example, Utpaladeva and his commentator Abhinavagupta’s interpretation of Bhartṛhari become widely enough known so as to have exerted a likely influence on the commentator on the Vākyapadīya (hereafter VP), Helārāja, with respect to the doctrine of adhyāsa (superimposition) (Rastogi 2018, p. 7).

This is not to discount or downplay Bhartṛhari’s very real and direct influence on Utpaladeva’s doctrine. There is often significant overlap between Bhartṛhari, Dharmakīrti, and Utpaladeva in terms of vocabulary. And while their respective doctrines may vary significantly with respect to the precise meaning of that vocabulary, Bhartṛhari and Utpaladeva are often closely aligned. One strong example in recent scholarship is Ferrante’s treatment of svasaṃvedana, a cognition’s quality of self-awareness. There, Ferrante argues, “the main contention of this article is that Utpaladeva and Abhinavagupta offer a peculiar and restrictive interpretation of svasaṃvedana in order to mount an attack on the Buddhist idea of nairātmya” (Ferrante 2017, p. 3). And, he concludes, “Bhartṛhari is a documented source of inspiration, especially in relation to the notion that a cognition is always restricted to itself and is never the content of another.” (p. 23). Ferrante demonstrates a strong overall influence from the VP to Utpaladeva and Abhinavagupta’s conception of svasaṃvedana.

These few examples from recent scholarship hopefully suffice to give a basic background to the close interweaving of Bhartṛhari’s thought with Utpaladeva’s, the former strongly influencing the latter, even while the latter innovates upon the former. The idea of svasaṃvedana, as it happens, closes out the SS with Utpaladeva quoting it directly from Dharmakīrti only to then juxtapose it with the corrected Śaiva form, where it constitutes a universal, transcendental subjectivity. I hope to look at this in detail in a future work, but will not take it up here.

Relative to Utpaladeva’s Pratyabhijñā (ĪPK and ĪPKv), there is an even more explicit emphasis on the grammaticality of existence in his SS. Such grammaticality of existence has rightly been compared to, among other things, Peircean pragmatic semiotics (Lawrence 2014, 2018). Incipient interpretability is present and real even in what is common-sensically called perception, that is, perceptual as opposed to conceptual cognition. There is in the SS accordingly a greater emphasis upon the singleness and fundamental unity of sambandhatvā (relatedness), the doctrine that the relation–principle is in fact one reality, and various everyday expressions of relation are just facets of that unity. Going further even than the realist, Nyāya-Vaiśeṣika categorical definition of a sambandha: that it be just one unitary involvement between only two relata, Utpaladeva not only preserves the realists’ requirement that a relation have two terms (Torella [1994] 2013, p. 96 n. 23), but here in the SS posits an underlying unity at the root of all, apparently separate, relations.

This unity within apparent diversity is, in due course, revealed as a feature of the unified though still variegated nature of awareness itself. Pure divine awareness, available even within the bound, limited perceiver, has the power to take in a multifaceted reality in a single act of perception. Utpaladeva’s two shifts in the SS—the increasing semiotic emphasis and increasing unity-emphasis—are not at all unrelated. A shift toward a more prolonged examination of semiosis reflects a tacit acceptance of a Bhartṛharian doctrine. This is of course none other than the doctrine, made explicit in SSam 3–5, that there is an underlying and unified reality to relation that is not directly expressible, even while indirectly encompassed by various linguistic expressions (Houben 1995, p. 170). All such expressions tend to pass for relation colloquially, though imprecisely. They, at best, approximate an underlying reality of the relatedness that Utpaladeva, following Bhartṛhari, has in mind. Relatedness, or relation in its absolute sense, is what orders phenomena into the distinct and unmuddled. Relation is thus not something objectifiable, but rather is indicative of phenomena as having the nature of vimarśa (reflective awareness), a core feature of divine subjectivity itself. What passes for relation are only its various manifestations, wherein subjective awareness interfaces with the manifest world.

Utpaladeva follows the Bhartṛharian doctrine of śabdabhāvanā, borrowing the grammarian’s own example of the person not consciously aware of wayside objects while on a journey but able to recall and name them later. This example is borrowed from the auto commentary to the VP. Depending on whether one accepts Iyer’s or Rau’s verse numbering, it may be VP 1.123 (Iyer 1965) or VP 1.131 (Torella [1994] 2013, p. 125 n. 42), respectively. In Utpaladeva’s reading, it is the relatedness of objects that orders them into distinct and unmuddled sensory experience, even at the perceptual level.3 Utpaladeva thereby defends the idea of linguistic articulation being incipiently present in perceptual cognition. It is as though the SS serves as a counterpoint to Utpaladeva’s commentary on the ŚD. In that text, he does not venture to contradict his master Somānanda’s adversarial disposition towards the grammarians (Nemec 2011), whereas in the SS he continues on his own initiative to deploy Bhartṛharian ideas even further beyond the reach of his Pratyabhijñā.4 This shift in substance of Utpaladeva’s argument, toward a further embrace of the Bhartṛharian, is accompanied by some significant rhetorical choices. Of particular interest is the quotation from Dharmakīrti’s apoha section of the Pramāṇavārttika (hereafter PV), which Utpaladeva deploys to attack Dharmakīrti’s nominalism directly.5 Thus an explicit targeting of Dharmakīrti’s nominalism occurs alongside an explicit deployment of Bhartṛhari’s semiosis. For though Bhartṛhari is never named nor quoted, Utpaladeva’s borrowed example above would surely have been instantly recognizable to his learned audience.

Utpaladeva pursues a worldview that is thoroughly semiotic, tinged with an emphasis on all reality as manifestation that he shares with the Dharmakīrtians; a kind of neo-Bhartṛharianism. The SS is an especially vivid expression of this. Its marked innovation upon the theory of Bhartṛhari concerns the emphasis Utpaladeva places upon perceptual acts of superimposition. What superimposition means, in this context, is that the semiotic configuration of a perception reveals or conceals the relations within that perception. Perceptions are, we might say, in a very real sense parsed by the subject in one way or another, either rendering their existing semiotic relations visible or not. In this sense, despite a trenchant refutation of Buddhist nominalism, Utpaladeva’s objective idealist system is a synthesis between Dharmakīrtian and Bhartṛharian innovations. On the one hand, Utpaladeva accepts the Dharmakīrtian idealist doctrine of reality as pure manifestation (ābhāsa, avabhāsa, pratibhāsa), but on the other hand, observes that such pure manifestation has an objective nature precisely because it is shot through with Bhartṛharian semiosis. All is manifestation, but meaning resides at every level of it, including purportedly pure sensation. In sum, the SS is neo-Bhartṛharian chiefly in that its theory of relation borrows the Buddhist emphasis on ābhāsa, etc., and in so doing places a greater emphasis than Bhartṛhari himself of the role of superimposition, which includes perception manifesting according to the way it is parsed by the observer.

2.2. Anti-Nominalism

In a departure from previous works, the SS is exhaustive in its exploration of Sanskrit grammatical examples, which advance a thoroughly anti-nominalist semiotic, establishing continuity-in-change where the Dharmakīrtians emphasize radical discontinuity. Dharmakīrti’s sensory world is one bereft of mind-independent relations, universals, or properties. Utpaladeva’s corrected worldview is one in which relations are intrinsically present wherever there is content available to experience, whether such relations are parsed by awareness or not. Like the Dharmakīrtian vision, for Utpaladeva, all experience is manifestation. Unlike it, such manifestation has the quality of either concealing or revealing the fact of its relatedness. Dharmakīrti’s world is radically discontinuous. In fact, its proponents might well point to this discontinuity between discrete moments of perception as a source of its empirical strength. Utpaladeva’s world is, by contrast, a world of change according to subjective perception, yet at the same time a world of continuity according to certain fixed, intersubjective features of a consensus reality. Utpaladeva’s world, then, affirms continuity-in-change as an unbroken whole. The SS thus mounts a distinctive attack on nominalism by describing (1) How relation really abides in the world of manifestation, despite being not always readily apparent in everyday grammar; (2) How such relation supports a Śaiva tripartite rather than Dharmakīrtian binary reality, with degrees of error rather than absolute truth and error; (3) How proper perception, therefore, derives from taking individual objects from out of their contextual whole, rather than in true isolation. Finally, I also briefly note how (4) Utpaladeva in the SS attacks a characteristically Dharmakīrtian form of apoha nominalism, itself a marked improvement on that of Dignāga. The following sections take these up each in turn.

2.2.1. The SS’s Doctrine of a Relation within a Single Word

There are some immediately obvious differences in how the later and earlier Utpaladeva approach the topic of relations. The section of The Pratyabhijñā that is most directly complementary to the SS, the āhnika on bhedābheda, together with its autocommentary (ĪPK and ĪPKv 2.2), deploys just two grammatical stock examples as the semiotic portion of its defense of relations. These are the genitive relation rājñaḥ puruṣa (The King’s man) (ĪPKv 2.2.4), which expresses a past and ongoing relation, that of servitude, and the concept of the kārakas (ĪPK and ĪPKv 2.2.6), those action-factors that give fine articulation to the small actions subordinate to a complex action, creation, or process, such as cooking.6 The latter expresses a present and ongoing action. I have touched on Utpaladeva’s deployment of the kārakas in a previous work (MacCracken 2017).

Despite the distinction of being based on past vs. present action, these two examples from the Pratyabhijñā are alike enough, as examples of relation that, when expressed linguistically, occur at the level of the uncompounded phrase and complete sentence, respectively, as opposed to a compounded phrase or even a single word. Utpaladeva’s comprehensive consideration in the SS extends to such sentential and uncompounded phrasal relations, but also extends further still, to those relations within single utterances: Sanskrit compound phrases or even single words (śabdas), such as the compounded version of rājñaḥ puruṣa, rājapuruṣa. The example of the uncompounded phrase rājñaḥ puruṣa is also used in the Pratyabhijñā to show how the man and the king he serves are both at once distinct empirical objects and are united in awareness. In the SS, Utpaladeva further emphasizes that from a vyavahāra standpoint—vyavahāra here meaning language usage expressive of consensus, transactional reality—it can also be the case that the sense of an ultimately existing relation vanishes. The compound phrase rājapuruṣa expresses the thought of the king’s man as a unified idea (very roughly in English, “kingsman,” if one likes), concealing the fact of any relation and emphasizing just the man.7 Relation may abide in the ultimate sense, but its concealment often predominates in everyday language and thought. Interestingly, this variation in the manifestation of relation correlates to some distinct experiential levels of reality that, true to their Śaiva panentheistic origins, have a tripartite rather than a dualistic character, with parā (absolute), aparā (relative) and parāpara (intermediate) levels.

2.2.2. The SS’s Levels of Truth

What are the discernible levels of relation in the SS? Despite relation’s indispensability to the act of manifest creation, it is not totally omnipresent or uniformly perceivable. Characteristic of a panentheistic system, there is an absolute and transcendental sense in which relation, or relationality, cannot be said to exist at all. There is an intermediate sense in which creation proliferates via a relatedness that both unites and divides. Finally, there is the appearance (if only a specious appearance) of the purely objective. There, those things that conventionally go by the name ‘relation’ (grammatical upakārya-upakāraka ‘aided-and-aiding’ relations) may reveal or conceal themselves. Those three levels of relatedness then stand as follows: Transcendental subjectivity has no relatedness itself. Yet, that subjectivity encompasses the act of phenomenal experience, which is fundamentally built on relatedness. But, distinct objects of experience have what we might call a purely objective mode, in which their relatedness is entirely obscured.

In the ultimate sense, sovereign and omnipresent awareness as such, characterized in the benedictory verses as viśvātma, prakaśa parameśvara, etc., has no intrinsic relationality, even if relatedness is one of its distinctive powers.8 Such awareness extends beyond and transcends subjective and objective semiotic categories alike. In Sanskrit grammatical terms, this awareness is a pure agent and is not a relatum in any grammatical upakāryopakāraka process.9 For this reason, relation is inherently a feature only of the world of pradhāna, synonymous in Kashmiri Śaiva vocabulary with bound experience, in the cosmogonic sense.10 It is relation that coordinates experience of the vyavahāra world. And, in the experiential sense, those proceeding in a world of relation are bound beings who, nonetheless, have greater purchase on their bound condition by embracing relatedness as real and thus a pragmatically generative principle to understand and use.11

Standing, as it were, at the opposite grammatical end from the particular is the entire sentence relationship expressed by the vibhaktis: the nominative, accusative, instrumental and so on. Their grammatical complements are the action–factors; the kārakas (ĪPK and ĪPKv 2.2.6, SS passim) It is these that Dharmakīrti reductively attacks as only the mind’s invention. But for Utpaladeva as well as for the grammarians, kārakas are in fact the real-life semiosis embedded in any given action. In his benedictory verses, Utpaladeva forcefully attacks nominalism in his mention of the Lord’s unfolding beginning with (ādi) the semiotic categories. I have touched on this before (MacCracken 2017, 2021), so will not recount it in detail here, other than to say this is an anti-nominalist move in that it reverses the Dharmakīrtian logical order of cognition characterized as first pure experience, followed by conceptual elaboration into universals, properties, etc., and only then, further elaboration into bhāvabheda relations of nominative, accusative, instrumental, etc.

The Dharmakīrtian account, which interprets the Buddhist absolute/relative distinction in terms of absolute pure sensation vs. relative inference, tends to bifurcate the world. Accordingly, its account likewise divides the world as above, only with pure sensation as the sense in which relation does not exist, soteriologically aligned with awakening, while the world of relation is only relative, provisional, and linguistic, tied to inference. This makes its way into Utpaladeva’s pūrvapakṣa (opponent) in the SS. Dharmakīrtians are portrayed as explaining phenomena like the avayavin (part–possessor) as an illusion characteristic of saṃvṛtisaṃjñayā (relative knowledge), but without further accounting for degrees of relative knowledge.12

Utpaladeva posits a third reality, however. As part of the vyavahāra world, it can be said that relations are fully present even in a single particular, such as the blue lotus. The nīlam utpalam, as he puts it, is a lotus distinct from other lotuses simply via the copula, asti, that specifies it as a lotus distinct from those of other colors.13 It is a Bhartṛharian doctrine that a sentence is defined by parts of speech grouping around a single main verb. Here, Utpaladeva emphasizes the status of existence itself, expressed by the copula, as a type of action. This relationality between the subject and predicate succinctly conveys the idea that what has been perceptually distinguished—recognized—is always already in relationship, by definition. Such an adjectival relationship in a single object functions not unlike a genitive relationship between persons, like the rājñaḥ puruṣa example, the difference being that the king’s servant is related by the action of servitude, whereas in the lotus the kriyā, the creative action (ultimately Śiva’s) is there at the immutable level of the object, regarded as a bundle of properties. The relation is quite simply a certain type of being-as. Just as in grammar, so awareness also, in a very real sense, parses experience based on the action of the object, down to its very fundamental quality of existing: a property, but also, in a real sense, an action.

In the course of his defense of relation, Utpaladeva reiterates his prior argument from the Pratyabhijñā that Śiva, “another” perceiver, is needed to relate cognitions otherwise incapable of connection.14 The existence of relations thus implies the presence of Śiva via his power of vimarśa. But the extended defense of relation as a bhedābheda reality is, after all, a pragmatic statement. Utpaladeva enjoins upon his audience an affirmation of the effectiveness of recognizing relation for the real that it is. The SS is, then, very much concerned with making sense of the manifest world of experience. Utpaladeva devotes a substantial amount of his attention to exploring how perception superimposes itself onto this or that unitary cognition. The way that individuals perceive may parse an idea in this way or that, and its grammatical expression may be subject to certain conventions, but one should bear in mind that grammatical relations are descriptive of real features of consensus reality, and it is their existence that makes experience ordered and intelligible: a distinct web of one-to-one involvements as opposed to a muddled mass of indistinct homogeneity. Such ordered two-term relations can be expressed grammatically in a number of different ways. Utpaladeva cites a few examples not characteristic of his prior works, including some laconic but interesting statements on the pragmatically implied presence of an object, and devotes extended attention to coordinative compounds (passim). The former example alludes to the relatedness between present and absent objects. The latter alludes to the capacity of awareness to group distinct subcategories into a single category, and so on.

2.2.3. Pragmatic and Unitary Superimposition

An emphatic doctrine of the SS is that objects manifest in the form of the perceptual cognition apprehending them, and this can come through in a number of ways. Some ways convey a highly pragmatic sensibility concerning how perceptual judgments are formed, while others emphasize the unitary nature of a perceptual cognition.

To begin with, pseudo-perceptions, objects simply pictured in the mind’s eye, retain a perceptual character. It is possible to picture them without any conceptual elaboration.15 Committing errors of substitution is a similar process. One can, as an instant reflex, though mistakenly, see silver in a seashell without reasoning through the process. The mistaken perception is simply that, with no error in conceptual cognition involved.16 Further, one has the perceptual cognition of absence of a pot in its proper place, etc., via a contrastive relation to its antarnīta (implied) sense, the lingering transverbal impression, of its presence.17 Or in the case of a cognition as expressed in a phrase like, “The king’s servant, and the Brahman’s,” the presence of the second servant is implied pragmatically.18 A great deal of Dhamakīrti’s philosophy also famously trades on a set of pragmatic considerations around how perception is formed, according to the proclivities of the individual. Here, I have in mind the well-traveled gruesome example of how the body of a dead woman means different things to “a dog, a libertine, and a yogin” (Dunne 2004, p. 184). The difference here is that, in Utpaladeva’s pragmatism, it is nonsensical for the Dharmakīrtians to assign to the relations implied by such pragmatic considerations the status of total illusion, such as the stock example of the rope mistaken for a snake. Such pragmatic considerations produce determinate results not later contradicted, and thus aid in accurate judgments and predictions, etc., where a true illusion would instead simply vanish completely and irreversibly in the light of truth. Dharmakīrti’s binary worldview of perception (real in the absolute sense) vs. inference (real only in a relative sense) ought also to admit a third category, that of immaterial though real predictive features of consensus experience, relatedness foremost among them.

Perceptual cognitions, on the other hand, retain the form superimposed onto them by perception in the everyday transactional sense. This semiotic difference in how experience is parsed in perception is further crystallized when encoded into spoken language. A vivid example of the former doctrine is the difference in how Utpaladeva recounts the uncompounded phrase rājñaḥ puruṣa vs. its compounded equivalent rājapuruṣa, where relation in the vyavahāra sense is present in the former, though not in the latter.19 Though both have the same referent, presumably the one is a perception that emphasizes the relationship per se, whereas the other perception emphasizes the whole and single man who bears that relationship.

Also a type of superimposition may be seen in the vyavahāra sense in the case of Sanskrit compounds. Items of perception that may be grouped together into a single grammatical remainder take the form of a single cognition, as expressed in coordinative compounds. In these cases, again, what matters is the pragmatic understanding of the total meaning. A pragmatic understanding—that is, one that is contextually semiotic—should be understood as something that grammar both encodes and is derived from, and to which spoken grammar is, in that sense, subordinate. The predominance of such pragmatic understanding may be seen in, for example, its overriding of grammatical eccentricity. For example, grammatical rules may specify gender and a number or, in some cases, an eccentric word whose number and gender both are at variance with the meaning expressed. The unified shape of a cognition, in other words, matters in the ultimate sense whereas considerations of number, gender, and even the choice of word expressing a concept are all ultimately arbitrary.20

But, conversely, all such grammatical features do nothing at all to detract from the underlying meaning. Meaning abides in a sense that is untouched and unaffected by either grammatical convention on a surface level, or by the tendency to parse cognition this way or that way according to either one’s long-term proclivities or present interests.

Accordingly, Utpaladeva’s latter points on the arbitrariness of language may at first recall Dharmakīrti’s position. But they are in fact a significant modification. The difference is that Utpaladeva is pointing out an ordered and determinate semiosis within cognition involving real existing universals, relations, etc., which language approximates and onto which sometimes arbitrary grammatical rules are superimposed. Moreover, from one perceiver to another, different relations may be visible or obscured due to, so to speak, the shape of one’s very thoughts. Dharmakīrti’s argument takes a different direction, finding in the arbitrariness of relation between word and meaning ample evidence for his doctrine of perception and language as fundamentally separate. For Utpaladeva, perception itself is thoroughly semiotic with Sanskrit grammar following closely after it. This is as against Dharmakīrti, for whom Sanskrit grammar following from perception is taken as evidence of perception’s wholly non-semiotic nature. On Dharmakīrti’s nominalism, one additional point is also worth raising concerning Utpaladeva’s choice of words in constructing his Dharmakīrtian pūrvapakṣa.

2.2.4. Dharmakīrti’s Identity of Vyāvṛtta and Vyāvṛtti, and Dharmakīrtian Nominalism

A small though intriguing point is Utpaladeva’s word choice in reconstructing his opponents’ view, where an everyday object of perception is characterized in explicitly apoha terms, as a vyāvṛttam (excluded thing).21 That is, the identification of a particular occurs through the famous apoha nominalist process of negation, wherein an object is distinguished from other objects and takes its svalakṣaṇa (defining characteristic) on the basis of its being excluded from the class of things it is not (e.g., the cow being classified as that which is not-not-cow, etc.). Although vyāvṛtta is part of a Buddhist vocabulary predating Dharmakīrti, it might also be remarked that a specifically Dharmakīrtian sense of apoha, as opposed to Dignāga’s apoha, turns on the doctrine of the precise relationship between vyāvṛtta and vyāvṛtti: the excluded thing in the apoha nominalist scheme, and the act of exclusion itself. Utpaladeva rightly does not burden his Śaiva audience with a detailed history of the evolution of Buddhist apoha, but only its Dharmakīrtian expression. Still, intentionally or otherwise, his mention of vyāvṛttam invokes a characteristic difference between Dharmakīrti and his predecessor.

To recount only briefly the history of apoha’s development, which I drawn from (Taber and Kataoka 2017), Kumārila rightly becomes a champion of Mimamsa-based arguments, by offering an effective criticism of Dignāga’s version of apoha. Kumārila’s claim, explained by Taber and Kataoka, makes an analytical division as to whether words for Dignāga indicate apohas themselves, or things possessing apoha, with neither being logically acceptable to the Mīmāṃsā. In the former case, Kumārila reasons, then two words such as ‘blue’ and ‘lotus’ could not be qualifier and qualificand. The two negatives have no way to intersect. On the other hand, as Taber and Kataoka explain,

if a word is taken to refer to the thing possessing the apoha, analogous to a thing possessing a universal, then a problem analogous to the one raised by Dignāga for a word indicating a thing possessing a universal would arise. Namely, it would not indicate its meaning directly but “dependently,” that is to say, it would refer to the particular insofar as it possesses just that apoha without any implication of others. Thus, one word would not “cover” the meaning of another, making coreference impossible.

(Taber and Kataoka 2017, p. 262)

As the authors continue,

For Dharmakīrti the key to the solution of the problems of coreferentiality and qualification is seeing that there is no real distinction between exclusion (vyāvṛtti) and the thing that is excluded (vyāvṛtta); their distinction is based merely on convention. If the excluded thing were something different from the exclusion, then it would be among the things excluded—would be a non-cow. Given that there is just the exclusion and not an excluded thing different from it, one may nevertheless choose to refer to it in isolation, as a unique exclusion set off from other exclusions, or as an exclusion coexisting with other exclusions. In the former case, it appears as a property; in the latter, as a property-bearer. Dharmakīrti stresses in this connection that words can mean whatever we want them to mean—under the constraint of course that they refer to exclusions.

(Taber and Kataoka 2017, p. 264 emphasis mine)

Though these arguments are intricate, and the whole of the above cited chapter is highly recommended for further context, hopefully the above quotes suffice here to demonstrate how the Dharmakīrtian version of apoha turns on the nondifference between exclusion and thing-excluded, and their distinction in terms of mere convention.

Again, Utpaladeva does not thematize this explicit development from the apoha of Dignāga to that of Dharmakīrti, nor would it make rhetorical sense for him to do so. But it is nonetheless evident that Utpaladeva means to account for apparent change in manifestation of relation from one facet of experience to the next, or one perceiver to another, while reining in the suggestion that complete and total arbitrariness of word-meaning relation prevails. For Utpaladeva, relations have real existence, though the character of different cognitions may disguise the fact in vyavahāra terms, so that convention does determine the apprehension or non-apprehension of relations thereby. This can happen not just through the accident of grammatical eccentricity, but through a process not unlike what Dharmakīrti has in mind above. One and the same object of experience may at times appear as something more like a property, and at times something more like a property-bearer. The difference is, of course, that for Utpaladeva relations have real determinate and predictive existence, so that they can help to discern individual manifestations against the background of manifestation as such, as opposed to Dharmakīrti’s worldview, wherein objects are in themselves nondifferent from the negations distinguishing them, yet eternally distinct and bound within their own confines. Utpaladeva thus undermines a specifically Dharmakīrtian form of nominalism by pointing out a semiosis occurring logically prior to grammatical formulations, abiding within the shape of perception and mental image alike, and finally encoded, often imprecisely, in language.

2.3. Continuity-in-Change in the SS

What does the body of doctrine expressed in the SS, then, amount to in terms of continuity and change, and therefore contribute to our understanding of Utpaladeva’s view of the same? In furthering the defense of vyavahāra experiences as bhedābheda, an indissoluble continuum of continuity-in-change, it certainly develops upon The Pratyabhijñā. However, it does so with an even more explicit emphasis on the Bhartṛharian, focusing on a range of supporting grammatical examples including individual phrases, compounds, etc. And in a neo-Bhartṛharian move, the Dharmakīrtian idealist emphasis on experience as manifestation is accepted, though modified to an objective idealism wherein relations not only exist but are fundamental to the very act of perception, whether everyday language reveals this fact or obscures it.

What of Dharmakīrti’s vicious regress argument? Relations are conventionally understood as either temporal or spatial. Indeed, they are analyzed by Dharmakīrti this way in his SP.22 They cannot be temporal for Dharmakīrti. Although Buddhists affirm causal processes as real, a causal process does not fit the Nyāya-Vaiśeṣika definition of a relation as one involvement between two relata, as already mentioned above (p. 5, Section 2.1). A causal process: be it combination, separation, phase-transition, etc., occurs in time and so at no single point in time is there both a cause and effect, and hence no two relata to relate. But neither can relations be spatial. Nor again can relations be even conceptually real for Dharmakīrti. And this is where Dharmakīrti’s regress comes into play. Objects cannot be related by a real relation, for that selfsame familiar problem Bradley raises. A vicious regress of proliferating relations would be required to connect the whole thing together.23

This argument is in fact quite effective against a naïve realist sense of relation as existents. However, Utpaladeva’s analysis in the SS seeks to demonstrate that no such proliferation is required due to relation’s fundamental nature of unity-in-diversity, and continuity-in-change. In the text, on the one hand, a distinctive emphasis on Sanskrit compounds explores at length how their relations remain clear and express coherent thoughts via the specificity of strictly two-member relations. For example, a king’s relation to his military units: horses, chariots, troops, etc. may in fact contain many distinct relations: Horses, chariots, and troops are all related as common subjects, a single category. Each horse, each chariot, each troop are all grouped together, a subcategory. Each relates, as a single subcategory and collective category, back to the one king.24 Each distinct two-term relation allows one to make sense of thought, and to express it in language. What we mean by the various relations is just one thing common to them all, and it metaphysically relates back to the power of creative action, the kriyāśakti, of being in the king’s active service.

Different perceivers might also vary in how they apprehend a single object. One might conceive the king’s servant in a thought expressed by the uncompounded phrase “The man of the king” which clearly emphasizes his genitive property, his “of-the-king”-ness. Or again, one might conceive him with more the shape of thought expressed “Kingsman,” and thus think of him in terms of a property bearer. The outcome is that, not unlike the specifically Dharmakīrtian variability described above (Section 2.2.4), a man may be described in terms of that to which he belongs, or that which belongs to him. The difference is, of course, that there is a subjective variation rather than total arbitrariness to consensus reality as a product of a semiotic rather than a nominalistic conception of how perception functions.

Moreover, relatedness is fully present at the level even of the single object. This is reflected in everyday language in the case of an uncompounded genitive phrase where the relation reveals itself, as opposed to the self-obscuration it has in the case of the compounded phrase. When we describe a lotus as blue, we are placing it in implicit relation with innumerable lotuses of other colors (or perhaps, blue things of other types), on account of its very fundamental action, a certain type of being-as. Its manifestation as a certain thing, we might say, is already a mode of a certain type of action, that of simply being a blue lotus. Continuity and change are not finally separable, but rather are aspects of the same process, variously perceived. Where the SSam of Bhartṛhari presents change and difference as an abstraction from out of a holistic continuity (Houben 1995), and the SP of Dharmakīrti reduces change to a discontinuous series of mental moments expressed in words that are totally arbitrary (Jha 1990), Utpaladeva’s SS establishes continuity-in-change as a continuum, modifying Bhartṛhari and repudiating Dharmakīrti. Utpaladeva affirms relationality, or the relation principle as a feature of a single transcendental subjectivity. Relation is, as such, not directly objectifiable and possessed of the power to conceal itself based on how it manifests to the individual perceiver. Relation is real, is not an existent, is omnipresent wheresoever there is dualistic worldly manifestation, and is possessed of the power of self-concealment.

3. Twenty-First Century Utpaladevean Relations

What constitutes the Dharmakīrtian is today relatively well-theorized, despite the scant likelihood of his radically discontinuous worldview finding wide favor in a contemporary setting. Despite what I argue is a much greater relevance to present-day concerns, what constitutes the Utpaladevean worldview is much less well-theorized. Recent scholarship has demonstrated that the more famous philosophy of Abhinavagupta, his gifts for stunningly vivid and lyrical communication of spiritual ideas notwithstanding, finds much of its original substance in Utpaladeva (Nemec 2012). Still more work remains to be done in connecting the original substance of the Utpaladevean with the present day, and this section outlines only a few possible steps in that direction.

Relation, as I have shown, is for Utpaladeva a real that, pace Dharmakīrti, encompasses unity-in-diversity, continuity-in-change. As a separate statement building upon Utpaladeva’s Pratyabhijñā, distinctive of the SS is its extended emphasis on the presence of relation down to the granularity of the single object or, as might be better said, the status of inherence as one type of relation.25 Distinctive also is its insistence that relatedness remains present in the absolute sense even while, in vyavahāra, consensus reality, the unifying shape of an individual’s particular cognition either reveals or effaces relations, as in a cognition shaped like either a compounded or uncompounded phrase. This variation in how relations are conceived or indeed whether they are apprehended as present at all is evidence not of the arbitrariness of relations as purely linguistic artifacts, as Dharmakīrti would have it. Nor again are relations metaphysical existents of precisely the naïve realist mechanistic model criticized by Dharmakīrti. Rather, relations have their quality of unity-in-diversity as a feature of the interaction between the individual perceiver and their environment. But as features of a consensus reality, relation is real in its variability, not unreal in total subjective arbitrariness. In what follows, I will not defend at length the way Utpaladeva’s theory of relations might fall amongst contemporary theories, but will make only a few suggestive comments.

3.1. Dharmakīrti’s Regress, or Bradley’s

To return to where we began, relation-theories throughout the 20th century remained a fiercely debated topic, on which turned much of the tension between realist and idealist ways of viewing the world. These achieve something of a culmination in the tension between Russell and Bradley. Closely tied to Bradley’s idealist tendency was his total repudiation of relations. Also in the 20th century, the discovery and publication of the Kashmiri Series of Texts and Studies introduced to the world outside of India the scriptural and exegetical writings of nondual Śaiva Tantrism or Kashmiri Śaivism. As scholarship in both these areas further develops, it is interesting to take stock of how Russell and Utpaladeva independently respond to skeptical denials of relation-theory.

Although an exemplar of the analytic tradition with which Dharmakīrti might find numerous affinities, Russell of course, nonetheless, comes down on the side opposed to a skeptical view of relations, universals, etc. He has, moreover, some points of overlap with Bhartṛhari. Russell thereby affirms relations in the world of objects, by following through on the logic suggested by everyday language. One late expression of this is his 1940 work, An Inquiry into Meaning and Truth, where he outlines his method as follows: “as opposed to traditional theory of knowledge, the method adopted differs chiefly in the importance attached to linguistic considerations” (Russell [1940] 1965, p. 9) Russell considers non-symmetrical relations such as “brother” particularly compelling examples of real relations that qualify the statement “the simplest propositions are relational.” (p. 33) This helps lead him to an observation Bhartṛhari would appreciate, that sentences must be understood in their unity, given the relations of words one to the other (p. 34). The main difference I would add, of course, is how English relates words together largely via word-order, while inflected languages like Sanskrit or Greek convey their relations by word-endings instead. Russell ends up delineating three separate types of relation: monadic (predicates), dyadic (e.g., ‘earlier’), and triadic (e.g., ‘between’) (pp. 42–43).

After a book-length treatment of linguistic considerations, Russell then turns in his short concluding chapter Language and Metaphysics to rule thusly on relations, “it seems that there is no escape from admitting relations as part of the non-linguistic constitution of the world” (Russell [1940] 1965, p. 325). And on universals:

I conclude, therefore, though with hesitation, that there are universals, and not merely general words. Similarity, at least, will have to be admitted; and in that case it seems hardly worthwhile to adopt elaborate devices for the exclusion of other universals.

(p. 327)

The above method of Russell’s, working through linguistic considerations and then deriving an anti-nominalist metaphysics from them, is arguably more productive than his earlier attempts to meet Bradley’s arguments as it were head-on, as I will now describe.

If the Russell of An Inquiry into Meaning and Truth stands well on his own, there has recently emerged reason to doubt that Russell’s victory over Bradley’s regress argument is quite as clear-cut as it might have once seemed. In Stewart Candlish’s excellent The Russell/Bradley Dispute, the textbook portrait of the tension between these two philosophers receives some important revisitation. After summarizing this textbook portrait of Bradley’s defeat and Russell’s victory, Candlish declares:

This is an historical question of some significance, for something like that summary has been important to analytic philosophy’s self-image, an image which depends upon contrast with that of a benighted and vanquished predecessor, idealism. And one of the battlegrounds on which it did indeed seem for many years that idealism had decisively lost was that of relations.

(Candlish 2009, p. 143)

Worth revisitation, for Candlish, is the frequent ambivalence in the mind of his opponents as to whether it is the internality or the unreality of relations that Bradley means to defend.

This ambivalence is compounded by the fact that Bradley himself seems not always at all clear on the matter, an obscurity further compounded by his different usages of the word ‘internal’ (p. 141 ff.). Indeed, I have avoided in the present work the use of the phrase “internal relations,” which is confusing enough between Russell and Bradley without adding what Indian Philosophy means by internal and external into the muddle. Interested readers should consult the Candlish volume as it is neither possible nor urgent for present purposes to summarize the matter succinctly. Suffice it to say that when Russell attacks Bradley’s regress, for Candlish, the reply misses the mark. Of Bradley’s regress, Candlish writes:

There is a standard reply to it which originates with Russell and is endorsed by Wollheim. It is that the regress is indeed endless but not vicious, being merely one of implication and not requiring the actual completion of an infinite series before anything can actually be related. (Thus ‘A and B are alike’ implies ‘A is like something which is like A’, which in turn implies ‘A is like something which is like something which is like B’ and so on ad infinitum but unworryingly.) This reply, if it is to be effective, must be based on the idea that the goal of the argument is to prove the internality rather than the unreality of relations…. If we keep it in mind that the question at issue is whether or not relations are real, we can see that the argument’s point is that an infinite series of actual objects is generated, not just an infinite series of possible names…

(Candlish 2009, p. 169)26

In other words, philosophers cannot lightly dismiss the fact that Bradley (and, I would also add, Dharmakīrti) are motivated by a total repudiation of real relations and not just the attempt to qualify them as one type or another. Accordingly, such skeptics argue against any vision of relations as a positive real, a positivity which, by this regress argument, would seem to lead to their endless proliferation. Dharmakīrti’s solution to the regress is of course a theory based around relation as a nominalist negative rather than a positive entity. For Utpaladeva, it would seem nonsensical to say that an infinite series of objects is generated in the relation process. Relations are, rather, evidence of a single and unbroken quality to awareness, its power of relatedness. What go by the name of individual relations do not proliferate, but are rather in a sense extracted from that holistic whole, as different facets of an unbroken continuum of meaning. This reasoning is typical of Utpaladeva’s distinctive method of transcendental argument and might seem exotic from the standpoint of the analytic philosophical tradition. Still, it is not a form of reasoning altogether foreign to present-day concerns. It bears a number of resemblances to another philosophical lineage, that of Peircean semiotics.

3.2. The SS and Present-Day Semiotics

Well-documented is the affinity between Utpaladeva’s (and Abhinvagupta’s) neo-Bhartṛharian conception of the world on the one hand, and a Peircean semiosis on the other. Distinct from Russell’s foregrounding of a linguistic method from which to draw metaphysical conclusions, Peirce’s semiotic late in his body of writings is not only a method. Rather, his semiotic comes to be widely encompassing of all methods, in a way that draws close to Bhartṛhari’s grammar. I want to emphasize two points here. First, Peirce eventually comes to see semiotics as both transcending and including logic and epistemology, in general accord with the way Utpaladeva considers inadequate and wrongheaded Dharmakīrti’s subordination of language to epistemology. Second, Peirce’s conception of scientific truth as intimately related to consensus is a process necessitating the interaction of self and other, and so draws close to the Kashmiri Śaiva method of meditating on the self-evidence of one’s own divine consciousness before guessing at its presence in other sentient beings. This Kashmiri Śaiva two-part method is the procedure of the ĪPK 1 and ĪPK 2, respectively. Thus the Utpaladevean corpus shares with Peirce an emphasis on truth as highly intersubjective, though Peircean-inspired philosophies up to the present moment differ from one another in their conception of pansemiosis.

First, concerning Peirce’s emphasis on the subordination of logic to semiosis, Edward C. Moore writes:

Peirce from the beginning conceived of logic as coming in its entirety within the scope of the general theory of signs, and all his work on logic was done within that framework. At first he conceived of logic as a branch of semeiotic (his preferred spelling). Later he distinguished a narrow and a broad sense of logic. In the broad sense logic was coextensive with semeiotic. Eventually he abandoned the narrow sense, and the comprehensive treatise on which he was working in the last decade of his life was entitled “A System of Logic, considered as Semeiotic”.

(Peirce 1982, p. xii)

I emphasize this point because it arcs close to Utpaladeva’s use of Bhartṛhari, where the logic of the day, Nyāya-Vaiśeṣika, is substantiated only via a thoroughly semiotic view of the world. Such a worldview also inverts the logic of a project of Dharmakīrti’s, considering logic as just one aspect of a world saturated in meaning, as opposed to epistemology as a means of dividing the world into the perceptual and the inferential.

What, secondly, does it really mean to have a thoroughly semiotic view of the world? In his excellent volume Cybersemiotics, which draws on a rich range of sources including recent developments in Peircean thought, Soren Brier draws the following distinction:

One version of pan-semiotics is constructivistic. Semiosis is everywhere, either because everything is semiosis in its nature, or because the only way we can know anything is through semiosis…. Reality is constructed by human societies living together in language…. Thus it is close to becoming a human-centered metaphysis (a subjective idealism) with no explicit idea of what nature could be in itself—or to put it in another way, what kind of external source there could be for the signs of nature….

The other version of pan-semiotics posits that signs are as real as atoms and energy and that the latter are also signs. Signs are the basic constituents of the world…. Sign processes are thus taken as intrinsic in nature. This interpretation finds support in quotes from Peirce like this: ‘Nominalism introduced the notion that consciousness, i.e., percepts, is not the real thing but only signs of the real thing. But…these signs are the very real thing. Reals are signs. To try to peel off signs & get down to the real thing is like trying to peel an onion and get down to the onion itself.’

(Brier 2008, pp. 357–58)

The striking element in Utpaladeva’s thought is that, I would argue, it is not entirely reducible to either of these, even if it leans heavily toward the latter. As far as the manifest relative world is concerned (i.e., wheresoever relations are operative), Utpaladeva is clear that they have real existence in terms of the determinate results to which their understanding gives rise. And at the same time, while it would be going too far to call Utpaladeva’s pansemiosis constructivistic, the world he describes is one in which different relations emphasize themselves from within a single thought from one person to another or even, from one thought to the next.

4. Conclusions

What I find fascinating in Utpaladeva’s writings is his commitment, both at once, to a type of idealism and to an affirmation of the phenomenal world. As Ratié demonstrates, there is an ethical component also to the Kashmiri Śaiva affirmation of other sentient beings (Ratié 2007). Such affirmation is not explicitly thematized by Utpaladeva in this text. However, as scholarship increasingly demonstrates that Utpaladeva originates the conceptual tools that Abhinavagupta uses as a faithful commentator (Nemec 2012; Torella [1994] 2013), I believe closer attention to Utpaladeva is warranted, including study of his affinity with Peircean semiosis, first initiated by Lawrence (Lawrence 2014, 2018), and a trend toward which I hope to contribute in future works. Though not devoted specifically to the question of other sentient beings, I find the SS interesting in its affirmation of the intrinsic relationality of all objects of perception: people, places, and things. Moreover, the appeal that it makes to its audience of practitioners and fellow philosophers—that belief in the reality of relatedness will positively affect outcomes in their lives—is an appeal whose basic supposition is that there are other beings out there that are worth the effort of liberation.