Beyond Time, Not Before Time: The Pratyabhijñā S'aiva Critique of Dharmakīrti on the Reality of Beginningless Conceptual Differentiation

This essay evaluates the philosophical merits of the Pratyabhijñā Śaiva critique of Dharmakīrti's stance that the judgment of sameness that constitutes a concept formed via exclusion (apoha) does not require ultimate grounding, and presents these Śaivas' account of the relationship between time, reality, and concepts. These Śaivas argue that time is the expression, as the subject/object pairs that define conventional worlds, of the non-dual differentiation inherent in ultimate reality—and use Dharmakīrti's own apoha theory to explain how this works.

The influential apoha (exclusion) theory of concept formation of the seventh-century Buddhist Dharmakīrti stands as a philosophically powerful articulation of how language could work in the absence of real universals. In brief, Dharmakīrti argues that concepts are constructed through a goaloriented process that delimits the content of an experience by ignoring whatever does not conform to one's conditioned expectations. There are no real similarities that ground this process. Rather, a concept is merely what's left over once one has glossed over enough of the differences between elements of one's awareness that one can (erroneously but pragmatically) judge the remainder to have the same effects as what one desires.¹

The question of whether or not Dharmakīrti's apoha theory can ultimately account for the judgment of sameness that is so crucial to the formation of a concept has formed the crux of some of the most vociferous critiques of this theory, both traditional and contemporary. Drawing on traditional Buddhist cosmologies, Dharmakīrti seems to address this thorny problem by positing the existence of a foundational error in the form of the subject/object structure of conventional experience. This error undergirds the process of concept formation, which cannot occur without the already given differentiation inherent in a world of conventional subjects and objects. Since this error is the expression of beginningless ignorance, it is therefore not subject to further scrutiny on Dharmakīrti's part. Relying on well-established and well-respected arguments to the effect that it does not make sense to inquire about the beginning of a causal sequence, Dharmakīrti seems to hold that designating subject/object structure as the expression of a beginningless karmic imprint (anādivāsanā) is sufficient to account for the differentiation within a moment of cognition in saṃsāra.²

While Dharmakīrti's articulation of the role of karmic imprints in the formation of a concept effectively addresses many of the objections surrounding the judgment of sameness, it also opens him up to at least two further lines of attack. The first concerns whether or not the Vijñānavādin ontology on which he ultimately relies can coherently support the claim that subject/object duality is non-conceptual. I would like to bracket this question for a moment and turn to a more fundamental critique, one to which Dharmakīrti is most likely subject regardless of how Vijñānavādins might respond: are beginningless karmic imprints sufficient to account for the mere fact of differentiated content in the conventional world?³ If subject/object structure is not conceptual, then Dharmakīrti must account for how differentiated content can emerge from a non-dual ground; if subject/object structure is conceptual, he must account for how a non-agentive, fully undifferentiated ultimate reality could have the kinds of limiting factors necessary to produce a concept according to apoha. In either model, then, the question becomes whether or not beginningless karmic imprints can do the work necessary for the creation of conceptual differentiation.

A highly sophisticated version of this critique plays out in the works of the tenthand eleventh-century Pratyabhijñā S'aiva tradition of Utpaladeva and Abhinavagupta. Although deeply influenced by the post-Dharmakīrtian Vijñānavādin tradition, these S'aivas follow a number of other Vedic traditions in heavily critiquing the Vijñānavādins' refusal to account for the origin of differentiation within the conventional world. The crux of this debate centers on whether or not it is possible to identify a root cause for the differentiation of the karmic imprints that themselves produce the differentiation experienced in the conventional world. Dharmakīrti's postulation that these karmic imprints are beginningless serves precisely to indicate that this question cannot be answered: since a beginningless process has no root cause, it is not subject to a circularity objection. These S'aivas reject this position and claim that the differentiation of karmic imprints does indeed require a basis. Even as they affirm that saṃsāra itself is beginningless—because saṃsāra is fundamentally formed by temporal processes and there is nothing that exists before time—they claim that the source of the beginningless time through which saṃsāra is experienced both can and must be accounted for by the grounding of the conventional world in an ultimate reality that is beyond time. For these S'aivas, rather than beginningless time being the necessary prerequisite for the manifest diversity of the conventional world, the non-dual differentiation inherent in ultimate reality is itself the source of time.

This essay will begin by examining the Pratyabhijñā S'aiva critique of Dharmakīrti's attempt to account for conventional differentiation by appealing to beginningless karmic imprints. Then, it will explore Abhinavagupta's own articulation of time as the expression of non-dual differentiation inherent in the ultimate itself. These S'aivas claim that the beginningless time through which saṃsāra manifests is the expression of the non-dual variegation inherent in ultimate consciousness. Driven by the freedom (svātantrya) of consciousness to will the creation of any possible world, time arises precisely as the conceptually constructed duality of subject and object. The cogency of the Pratyabhijñā critique of Dharmakīrti's use of beginningless karmic imprints raises doubts about whether or not Dharmakīrti's own version of apoha is the most philosophically sustainable. Rather than dismissing apoha itself as a viable theory, however, the Pratyabhijñā reformulation addresses the problem of the judgment of sameness by seeing beginningless conceptual differentiation as an expression of that which is beyond time.

Part I: Can Beginningless Karma Account for Conceptual Differentiation?

Dharmakīrti is hardly the only classical Indian thinker to propose that saṃsāra is beginningless. Moreover, the Pratyabhijñā S'aivas themselves affirm this position. Pinning down their precise objection, then, will require some clarity on broader ideas about beginninglessness in classical Indian philosophy. In an early article, Fernando Tola and Carmen Dragonetti document the ways in which many classical Indian traditions appeal to the idea that the worlds experienced within saṃsāra are beginningless (Tola and Dragonetti 1980). They identify the source of speculation about beginninglessness in post-Vedic traditions with the Upaniṣadic postulation of "two entities, Brahman and ātman, who exist in se et per se, without an element in them of relativity or conditionality" (1980, p. 2). As Tola and Dragonetti note, the type of beginninglessness attributed to these ultimate realities differs from the beginninglessness of saṃsāra in that "the Supreme Principle cannot be abolished, whilst the empirical reality can be abolished" (ibid.). Even in these early formulations, the specific type of beginninglessness manifested as saṃsāra stands in contrast to that of ultimate reality. The beginninglessness of ultimate reality is absolute, and equally entails an endlessness. The beginninglessness of saṃsāra, however, does not entail that saṃsāra, too, in all of its aspects is endless. If this were so, liberation would be impossible. Rather, the beginninglessness attributed to saṃsāra has to do with the impossibility of a causal process having an absolute beginning.

Steven Collins' exploration of the narrative function of nirvāṇa suggests a strong parallel in early Buddhist conceptions of the different types of beginninglessness represented by nirvāṇa and saṃsāra. As Collins indicates, the timelessness of nirvāṇa contrasts to time within the conventional world. Time in the conventional world is characterized as a concept referring to the mutual dependence of the various dharmas that form sequences: "The sequence of the three times is thus secondary, generated by and in the process by which conditioned Existents, which are also Conditioning Factors, give rise to more of the same" (Collins 2010, p. 35). Time is a concept abstracted from the apparent changes in sequences of dharmas. The particular relationship between time and the process of conditioning comes to the foreground in discussions of nirvāṇa because, as the ceasing of conditioning, nirvāṇa is also the end of time. As Collins explains, "The process of conditioning, and so of time, can self-destruct, so that time ceases to exist, at least for an individual" (2010, p. 38). In this way, nirvāṇa, unlike all conditioned phenomena, "has a relation to the past, but not to the future" (ibid.). This relation to the past is a conventional designation referring to the fact that a person constituted by a particular karmic stream seems to "nirvanize," to use Collins' verbal form, at a certain time, but "the temporal event denoted by such terms is not anything directly occurring in or to nirvana, but rather the ending-moment of the conditioned process" (ibid.). This supposedly temporal event foregrounds nirvāṇa's own atemporality through embodying the paradox of using finite verbal forms to refer to something to which the process of conditioning entailed by such verbs simply does not apply.

Along these lines, many Indian traditions accepted the two different types of beginninglessness that respectively characterize ultimate reality and saṃsāra: one that affirms a reality to which the categories of time simply do not apply, and one that affirms that it makes no sense to ask for the beginning of a temporal process. These two types of beginninglessness, then, distinguish the coherence of speaking of a reality that is beginningless because it is beyond time from the incoherence of asking what existed before time understood as manifest in causal sequences. The idea of a cause as an absolute beginning makes no sense because the specific characteristics of any cause come from the characteristics of the cause of which it was an effect. In contemporary parlance, the acknowledgment that a causal sequence cannot have an absolute beginning in time is the problem of the chicken and the egg. From within the perspective of the causal stream that constitutes the conventional world, asking for the beginning of saṃsāra is as futile as asking which came first: the chicken or the egg?

Dharmakīrti's account of the workings of the judgment of sameness that produces a concept uses the incoherence of asking for the first cause in a causal sequence as a bulwark against further questioning. For him, the universe of causally efficacious particulars, both subjective and objective, exists because of ignorance, expressed by and as the beginningless karmic imprints (anādivāsanā) that constitute the most basic structures of a sentient being's experience. Once causally efficacious unique particulars are in place, Dharmakīrti can appeal to a combination of subjective and objective factors to account for why the interaction between an embodied being, with particular sensory capacities, desires, goals, habits, and so forth, and a surrounding causal environment would lead to the production of concepts via exclusion.⁴ He thereby dismisses charges from his opponents that there is no way to account for the initial emergence of concepts. There is no way to point to the original chicken, either.

Although Dharmakīrti's move to align the most basic form of differentiation in the conventional world—the subject/object structure of a given moment of awareness—with beginningless ignorance plays on the philosophically well-founded rejection of the coherence of questions about the beginning of a causal process, he subtly shifts the location of the debate. In effect, Dharmakīrti appeals to the well-respected view that the karma of sentient beings is part of a beginningless temporal process in order to respond to an objection that, while closely related, is actually targeting a different type of differentiation. The incoherence of the chicken-and-egg question does not also mean that it is incoherent to ask what the nature of the chicken is such that it is capable of producing eggs. Depending on the direction and depth of one's analysis, one might find, for instance, that the chicken is composed of some kind of ultimate atemporal stuff, and that the categories of time and space that seem to define chickens and eggs are not absolute, but rather relative. The beginninglessness of causality, then, may itself presuppose a different kind of beginninglessness: a reality that is beyond time because time itself emerges from it. Both of these forms of beginninglessness may exist side by side precisely because they speak to different realities: the beginninglessness of causal processes is not the same thing as the atemporality of the ultimate.

The Pratyabhijñā concern with an original source of differentiation seeks to address why and how there could be differentiated stuff capable of entering into causal relations in the first place. This ends up being the question of how a single moment—any moment—could contain differentiated content at all. If this variety of content has no basis in ultimate reality, then, according to these S'aivas, it has no basis at all because an effect must have a real cause and something purely undifferentiated cannot cause differentiated effects. In short, these S'aivas will argue that one cannot appeal to the incoherence of something existing before time to claim that the question of the relationship between what is beyond time and the conventional world is itself nonsensical. If this critique is successful, then a key move in Dharmakīrti's defense of the judgment of sameness proves to be insufficient.

The Pratyabhijñā S'aivas' critique of the Vijñānavādin position that conventional differentiation is an expression of previous karmic imprints comes in the context of these S'aivas' exploration of what the nature of ultimate consciousness must be if it is to account for ordinary differentiated experiences. Utpaladeva summarizes his position in the Verses on the Recognition of the Lord (Īs'varapratyabhijñākārikā): "If manifestation is different [from objects] and non-different [in itself], then there is a conflation of the sensory object. Therefore, the object which is to be manifested has the nature of manifestation, and what is unmanifest is not established."⁵ This famous verse articulates a Pratyabhijñā position that will be widely cited and defended: there is nothing beyond the light of ultimate consciousness, which itself contains all possible differentiation in the form of preconceptual appearances (ābhāsas).

Isabelle Ratié (2010) describes how Utpaladeva's argument in this section proceeds by using two Buddhist accounts of how sensory objects could be related to consciousness to refute each other. Utpaladeva's Buddhist External Realist argues that external objects must be inferred to account for the variety of perceptions: "[Objection] Since consciousnesslight (bodhasya) being undifferentiated cannot be the cause of a multiform manifestation, all this various manifestation lacking in an apparent cause (ākasmika) leads to the inference of an external object [as its only possible cause]."⁶ Against this position, Utpaladeva will make use of Vijñānavādin arguments that the variety of karmic imprints (vāsanā) can serve this purpose, and so there's no need to infer external objects. However, the External Realist's objection anticipates this standard Vijñānavādin move and pushes further: "Not even a varied reawakening of the karmic residual traces can be taken to be the cause [of the multiform manifestations], for in that case a new question would arise: what is the cause of the variety of such a reawakening?"⁷ While the Vijñānavādin is correct to refute the position that external objects are necessary to account for the variety of experience, the debate does not stop there. Simply saying that the variety of karmic imprints is beginningless is not enough to respond to the External Realist critique. The cause of the variety of the karmic imprints themselves requires explanation. In line with the Vijñānavādin's own arguments, this cause can only be something internal to consciousness. To the extent that Vijñānavādins maintain that ultimate consciousness is purely undifferentiated, they cannot account for the basic structures of experience in the conventional world.

Ratié further notes that this is a common understanding of the Vijñānavādin position on how differentiated experiences manifest in the conventional world. Her quotation of the seventh-century Mīmāṃsaka Kumārila's presentation of this position in his Explanation of the Verses (S'lokavārttika) is particularly illuminating since Kumārila specifically identifies the beginningless flaw manifest within conventional experience with subject/object structure itself:

According to my [the Vijñānavādin's] doctrine, although the essence of cognition is in fact pure, nonetheless, in the endless (anādi) cycle of rebirths, because of a confusion due to the impregnations (vāsanā) that were born from previous cognitions [and] that are varied (citra), because they have causes that are [themselves] varied, the blue or [any other objective aspect taken on by the cognition,] stained by [the distinction between] the apprehended [object] and the apprehending [subject], arises while being seemingly differentiated in conformity [with its cause]; it does not require any other object [that would be external to the cognition]. And this relation of mutual causality (anyonyahetutā) between the cognition and the power ('sakti) [that constitutes the impregnation] is beginningless (anādika).⁸

Kumārila here accurately presents the Vijñānavādin denial that it is necessary to account for the initial cause of variegated karmic imprints. Differentiation does not actually have a root cause: cognitions contaminated by ignorance merely seem to be differentiated because they arise dependent on the beginningless error of subject/object duality. Moreover, on Kumārila's reading, the Vijñānavādin claims that the relationship between vāsanās and cognitions represents a logically acceptable case of mutual causality (anyonyahetutā), not a problematic circularity. This is the kind of causality typified in classical Indian philosophy by the example of the seed and the sprout. The precise reason why mutual causality is acceptable in this case is that the relationship between the vāsanā and the cognition, like the relationship between the seed and the sprout, is beginningless. To return again to the specific context of Dharmakīrti's apoha theory, if there's no problem with accounting for the basic differentiation of causally specific subjective and objective particulars, then there's no problem with basing the judgment of sameness that creates a concept via exclusion on this differentiation.

Although the justification for this position seemed to be well understood, the Vijñānavādin attempt to account for the differentiation manifest in the conventional world through an appeal to beginningless karmic imprints was still widely critiqued in medieval Indian philosophical circles. While Utpaladeva himself frames the critique as coming from a fellow Buddhist, Ratié points out that Naiyāyikas and Mīmāṃsakas are similarly skeptical (Ratié 2010, p. 460, nn. 65–66). While there is some variation among these critiques, they all revolve around the question of what the nature of vāsanās could be such that they could have the causal capacities attributed to them while still remaining nothing but consciousness. This question is particularly sticky since, according to these critiques, Vijñānavādins claim that ultimate consciousness is nothing but pure manifestation, and is therefore totally undifferentiated. Such an absolutely undifferentiated entity cannot itself directly be the cause of differentiation. If it were, everything should be caused all at the same time because a single undifferentiated entity would always have one and the same effect. However, if pure ultimate consciousness is not itself the cause of differentiated vāsanās, then the cause must either be something other than consciousness (in which case the Vijñānavādin would merely be another externalist), or something that is less than ultimately real, in which case the Vijñānavādin must explain how something unreal could be causally efficacious (pp. 456–458). The causal basis for the variety of experiences cannot come from something inessential to consciousness, whether that thing be an external object or an "adventitious defilement"⁹ that is ultimately left behind.

In sum, the basic form of the argument may be represented as follows: (1) We observe causally specific differentiation in the everyday world. (2) Something that is causally specific must be the effect of a specific real cause. (3) Each real cause produces only the effects that are in accord with its nature. (4) Such causes must be either internal or external to consciousness. (5) These causes cannot be external to consciousness because, per Vasubandhu and Dharmakīrti, external objects are irrelevant and logically incoherent. (6) These causes therefore must be internal to consciousness. (7) It is not the nature of something undifferentiated to produce different effects. For example, a cognition of blue has the causal capacity to produce only a subsequent cognition of blue, not a cognition of yellow. (8) Consciousness cannot be totally undifferentiated and produce different effects. (9) Consciousness must be inherently differentiated if it is to account for the differentiated awarenesses observed in the conventional world. Conclusion: Since there is no other viable candidate for the cause of this differentiation, the nature of reality is ultimate consciousness that inherently contains the capacity for the expression of all differentiated awarenesses.

A closer look at Dharmakīrti's responses to objections about the relationship between the natures things seem to have in the conventional world and what is ultimately real indicates that Dharmakīrti would not accept a key premise of these S'aivas' argument. For Dharmakīrti, things in the conventional world do not have to have real, ultimate causes. In a discussion of the judgment of sameness in Dharmakīrti's works, John Dunne proposes that Dharmakīrti's refusal to discuss why things have the nature they do is an intentional strategy aimed at avoiding irrelevant, incoherent, and counterproductive discussions. Dunne presents his favorable reading of Dharmakīrti's refusal to engage in metaphysical speculation:

If one is hoping for an ultimately defensible metaphysical reason, then Dharmakīrti's answer to the problem of sameness is dissatisfying. On the other hand, one might suppose that we are engaged in a frustrating and fruitless enterprise when we yearn to specify in precise terms the metaphysical warrant for our use of the term "fire." In that case, Dharmakīrti's answer is quite satisfactory, or perhaps even liberating.

(Dunne 2011, p. 99)¹⁰

This move parallels Dharmakīrti's striking refusal to disagree with an objector who, in response to Dharmakīrti's articulation of causality as the mark of what is ultimately real, protests that causality is merely conventional. Dharmakīrti responds: "Let it be so in the way as you have said."¹¹ Conventional causality is precisely that—conventional. Conventional causes can account for conventional effects; there does not need to be any deeper reality that grounds them.

Maintaining this position relies on a strict parameterization of conventional and ultimate reality. Conventional reality merely seems real; ultimate reality is actually real. Therefore, it does not matter that an undifferentiated ultimate reality is incapable of producing differentiated effects. Such differentiation is never actually produced; it merely seems phenomenally to exist to deluded sentient beings. The reason why it is not necessary to account for the ultimate causal basis of conventional differentiation is that this differentiation is beginningless. Like seeds that produce sprouts that produce seeds, all phenomena in the conventional world rely on their own previous causes and produce their own specific effects in a process without ultimate origin or grounding. The ability of sentient beings to judge that what are actually unique particulars share some subset of causal properties simply by ignoring what does not allow these beings to accomplish their goals—that is, the ability to perform a judgment of sameness that produces a concept via exclusion—is no more or less mysterious, no more or less demanding of an ultimate explanation than is any other ignorance-contaminated phenomenon in the conventional world.

Tom Tillemans expands Dunne's reading to further justify the wisdom of Dharmakīrti's refusal to answer questions about how to account for causal differentiation. Tillemans doubts whether or not this approach is ultimately successful, but, also following Dunne, he gives a positive evaluation of this failure: for these two scholars, Dharmakīrti's refusal to give an ultimate account of the grounding of concepts represents an "enlightened refusal" to engage in pointless ontological speculation (Tillemans 2011, p. 61). In the end, Tillemans sees Dharmakīrti as simply engaging in "a strategic refusal to justify metaphysically sameness that we do in fact recognize" (p. 60). In this way, according to Tillemans' terminology, Dharmakīrti offers an analysis of the correspondence between scheme and content that takes a certain judgment of sameness as a primitive fact that is not in need of further justification; he does not, however, provide an account that would fully justify this sameness (p. 60).

As Tillemans concludes, "The interesting feature of this version of bottom-up Apohavāda, if the theory is carried out consistently, would be Dharmakīrti's enlightened refusal to play a metaphysician's game that was best put aside" (2011, p. 61). Dunne cites Dharmakīrti's explicit denial of the legitimacy of questioning how particulars may be judged to produce the same effect in support of his and Tillemans' shared position that Dharmakīrti refuses to engage in metaphysical speculation. Dharmakīrti asserts: "Indeed, it is not correct (na… arhati) to question (paryanuyoga) the nature of things, as in 'Why does fire burn? Why is it hot, and water is not?' One should just ask this much, 'From what cause does a thing with this nature come?'"¹² Dharmakīrti's focus here on causality clearly indicates that he is comfortable providing a conventional explanation, but not an ultimate one. In this way, both Dunne and Tillemans affirm that Dharmakīrti's appeal to the causal capacities of an object, combined with subjective factors, cannot ultimately account for the judgment of sameness. Yet, for them, this supposed failure is not actually a failure, but a recognition of the inherent limitations of any attempt to ground the conventional in the ultimate.

In their assertion that manifest things must have real causes, the Pratyabhijñā S'aivas reject the idea that conventional reality is not grounded in ultimate reality. For them, the difference between various conventional realities and ultimate reality itself is not that they are real in different ways, but that they are real to different extents. Conventional realities are partial expressions carved out of ultimate reality through a process of exclusion driven by desire. The "reality" in both conventional and ultimate reality is the same reality. This contrasts strongly with Dharmakīrti's apparent position that ultimate reality does not have to causally ground conventional reality because conventional reality is a different sort of reality altogether.

Are these S'aivas, then, simply asking Dharmakīrti to do the impossible by specifying the beginning of beginningless saṃsāra? One crucial point here is that the Pratyabhijñā critique of beginningless karmic imprints does not explicitly concern the origin of saṃsāra itself. Both Dharmakīrti and these S'aivas will happily claim that saṃsāra writ large has no beginning; differentiation in the conventional world comes from karma, and it does not make sense to talk about karma having a specific beginning in space and time. The problem is not the beginninglessness of saṃsāra itself, but rather the connection between ultimate reality and conventional experience in any given moment: how does a differentiated experience arise out of a non-dual ground? These S'aivas are not asking for something that exists before time. Rather, having argued that time itself requires an explanation, they are asking what exists beyond. To sum up the S'aiva critique: based on the standards and assumptions widely shared by classical Indian philosophers, Dharmakīrti's use of beginningless karmic imprints is perfectly sufficient to account for the differences between various karmic streams within the conventional world. However, it is not sufficient to account for the mere fact that there is differentiated stuff capable of entering into causal relations. The Pratyabhijñā S'aivas offer a complex and distinctive solution to this problem: while they affirm that ultimate reality is beginningless in the sense that it is beyond time, they also claim that time itself has a "beginning" in the expression of the non-dual differentiation inherent in the ultimate itself. Time is nothing but the conceptual differentiation of subject and object that comes about through S'iva's use of the "chisel of exclusion," to use Abhinavagupta's provocative metaphor, to carve conventional worlds out of his own self.¹³

Part II: Abhinavagupta on Time, Conceptual Differentiation, and the Relationship between Conventional Worlds and Ultimate Non-duality

In his fascinating hybrid ritual-philosophical work, the Explanation of the Verses of the Mālinī [Tantra] (Mālinīs'lokavārttika),¹⁴ Abhinavagupta lays out the Pratyabhijñā understanding of the relationship between time, differentiation, concepts, and the expression of the ultimate as subject/objectstructured conventional worlds. According to Abhinavagupta, although the beginning of a temporal causal process within the conventional world cannot be identified, time itself has a "beginning" in the sense that it is the expression of an ultimate source that is beyond time. As an expression of ultimate reality, however, this "beginning" never quite loses its own connection to that which is beyond time—it is a beginningless beginning that emerges moment by moment from the play of non-dual consciousness. From the perspective of the ultimate, time never emerges for a first time. Within the conventional world, time is expressed as the delimitation, via exclusion, of causally specific subject/object pairs. Concepts are these delimitations of ultimate consciousness: they express slivers of that which is beyond time as the causally specific particulars whose interaction constitutes the beginningless time of saṃsāra. I will now explore these ideas in some depth.

The overall point of the Mālinī is to provide grounding and explanation for Abhinavagupta's position that the Trika¹⁵ revealed texts provide unique guidance because, as embodiments of S'iva's knowledge, they are expressions of ultimate reality itself. All other teachings develop as partial manifestations of them. As Abhinava explains:

The vast knowledge that is produced in the beginning (prāk) from the limitless reality (sadbhāva) that alone is identical with the world (bhāva) and that is the pervading nature (vaibhava) of the mass of the moon's (tad) rays has spread, [still] of the same nature (tādṛk), free from things to be accepted or shunned that are created by its own creative power (māyā), and diversified merely by its own expanding manifoldness of rays. This [knowledge] whose nature is articulation is the heart of the highest Lord.¹⁶

Abhinava goes on to emphasize S'iva's essential unity even though S'iva is thus the source of diversity. All revealed texts (s'āstras) contain an element of truth since they are aspects of S'iva's own self-revelation, but Abhinava's own Trika contains the most comprehensive articulation of truth possible.

An opponent, however, poses the following objection: "If [S'iva is] thus undivided and the self of the world, then how can there be—as the possibility of contraction is excluded—the riches of the S'āstras etc. which are rooted in the formation of differential thought?"¹⁷ Abhinava gives the standard Pratyabhijñā response that the unity of ultimate consciousness consists in the fact that the common element in all awareness is merely the capacity for manifestation, and then refers to the Pratyabhijñā position that the nature of consciousness is not only to reflect an object, but also to have a self-aware realization of this object.¹⁸ Bringing all this together, he claims that the process of the manifestation of an awareness begins within "the sphere of the experience of consciousness" and "becomes therefore (tadā) perceptible as soon as it appears as resting [inwardly] in this awareness; and it only later becomes a clearly perceptible [outward object]."¹⁹ Abhinavagupta then explains how S'iva places limits on his own knowledge through his own free will, thereby creating the differentiated forms of experience that constitute conventional worlds. As he sums up, "Even in this state, the conjunction and separation of constituent endless things become innumerable by combination (saṃdhāna) with the division of the earlier state. Only by virtue of these limiting adjuncts the various riches (vibhūti) of action and knowledge in the S'āstra give up the state of knowing consciousness [inwardly] to spread [in an objective form]."²⁰

These passages are exceedingly dense and present nearly all of the Pratyabhijñā's most important doctrines. In line with the dialogical format of most Indian philosophical texts, one could imagine many objections that Abhinava could pose in order to direct his further explanations. The objection he chooses to foreground is telling and points to the important role that time plays as the expression of the differentiation inherent in ultimate consciousness:

[Opp]: Then it would follow that divisions caused by space, time and [limited] power of action are not possible in this collection [of primary realities]. [A]: We certainly do agree, for there the tattva [called] 'time'²¹ is not known even by name. Although she pervades everything, the great goddess of time (mahākālī) does not manifest here. [Opp]: Then why do you accept the use of the words 'then', 'again', 'when', and 'afterwards' with reference to [something that is] undivided and complete in itself?²²

The opponent here rightly points out that Abhinavagupta's entire description of how ultimate consciousness produces limited objects of awareness is suffused with temporal language. Abhinava even begins his explanation of the unity of revealed texts by referring to "this vast knowledge that is produced in the beginning. …" If the ultimate is undivided, temporal distinctions have no place in relation to it. Therefore, there should be no "in the beginning" in relation to ultimate awareness.

Abhinava responds by not only granting that the ultimate is beyond time, but by affirming that since time is nothing but a particular expression of the ultimate, there is a sense in which time itself is beyond time! As he states: "We say that this is correct, but in reality these concepts of earlier and later do not exist for knowledge, even if the creation of tattvas has manifested perceptibly and time has unfolded."²³ The key point is that the limitations experienced within the conventional world, including temporal divisions, are both part of the ultimate because there is nothing that is not part of the ultimate, and unable to truly limit the ultimate because the ultimate exceeds any duality. Abhinava contends: "Therefore time is unable to cause differentiation in consciousness, nor is this time capable of becoming a differentiator [i.e. a differentiating quality] of the object of perception. For the universe does not exist outside of knowledge, otherwise it (tad) would not appear."²⁴ In short, the mere fact of differentiated appearances means that these appearances are somehow inherent in ultimate reality, even though they are experienced through limiting adjuncts such as time. Time itself cannot cause differences. Rather, it is an expression of the differentiation inherent in the ultimate.

Abhinava goes on to explicitly identify time with the expression of the differentiation of the ultimate: "It is only by causing the appearance of diversity that the Lord manifests time. This manifestation of diversity is termed 'the power of time'. Thus it is because of our (āsmākīnāt) accordance (anurodhatah.) with S'iva's (etat) power of time manifesting that qualifications [of time referred to by words] like 'then' etc. exist."²⁵ In this way, Abhinava closely links time to the non-dual differentiation that is inherent in ultimate reality. He continues to remind his opponent that, while their expression is constitutive of conventional worlds, both time and differentiation equally exist within the ultimate: "It is not [the case], that [time] does not exist in [S'iva] at all, [for] how can anything exist except in him. It could spread in another reality [and] would still be dependent on light."²⁶ Abhinava sums up the main point about time: "it has been said (kila) that as much as (yathā tathā) the highest Lord causes the construction of plurality to appear, indeed also time appears. But still he is never divided in the real sense. For, [as] he is consciousness, he simultaneously appears manifold [because of his autonomy]."²⁷ Time, then, is the expression of the inherent differentiation of the ultimate in terms of the experiences of limited subjects and objects.

As Utpaladeva and Abhinavagupta argue at length in their Pratyabhijñā corpus, this mutual exclusion of subject and object is itself conceptual: the mechanism by which S'iva creates conventional worlds is none other than apoha, and so concepts are inherently expressions of ultimate reality.²⁸ In the Mālinī, Abhinavagupta both references and reformulates the famous verse with which Dharmakīrti begins his exposition of apoha in the Autocommentary on the Explanation of Trustworthy Awareness. Dharmakīrti's verse, "All entities, due to being established in their own natures, naturally partake of exclusions from homogeneous as well as heterogeneous [things],"²⁹ aims to provide the fundamental rationale for why unique particulars may be judged to be the same despite there being no real similarities on which this judgment is based. Abhinavagupta uses the idea that "all things are established in their own natures," and are therefore inherently different from everything else, to indicate not that similarities are unreal, but rather that unity and difference are intertwined. After reminding his interlocutor that "even duality is not impossible in the non-dual reality,"³⁰ he states, "However, within difference, because all things are established in their own natures, I think that this word 'nondifferent' will become differentiated naturally."³¹ Here, Abhinava embeds Dharmakīrti's opening verse on apoha into his own discussion of the differentiation inherent in ultimate consciousness. Just as the presence of conceptual differentiation points back to a unifying source, the unifying source itself is not devoid of difference, but rather naturally expresses itself as the conceptual distinctions that shape conventional worlds.

Abhinava again aligns time with the appearance of the conventional world in the context of reversing the process of manifestation through the Krama ritual centered on Kālasaṃkarṣiṇī, the goddess who devours time.³² He describes the end state of this ritual as one in which "neither past nor future is divided from the present."³³ This process of moving outside the divisions of time results in the yogin becoming "one who moves in the void [of consciousness] (khecarah.)," who "has annihilated one's individual (nija) existence and relishes (carvaṇāṃ labhate) only the vibrant experience [of the nectar] of one's own immortality [i.e., the transcendence of time], [in which] flows an abundance (saṃdoha) of ambrosia that is the highest bliss."³⁴ However, Abhinava still emphasizes that the absence as well as the appearance of the world is not other than time:

For it is taught that time, which is the appearance of the world, is (yah sa) the vibration (saṃsphārah.) of the rays of one's own consciousness that is projecting (kalana) [the world]. The absence of the world is [also] it [i.e., time], it is nothing else.… But in the manner described [above], restraint, appearance, devouring etc. appear. And there is no other reality in the world than appearance in this way.³⁵

Here, the manipulation of time is aligned with whether or not one experiences oneself as existing as an independent subject in a world defined by the experience of time. This technique is effective precisely because time is both the expression of the conventional world and inherently present even in the still moment of realization that embodies that which is beyond conventional time.

In the context of his exposition of the nature of karma and the apparent limitations of agency expressed thereby, Abhinava draws on the idea that the "beginning" of conventional worlds is a beginning of time, not a beginning within time. He further uses the distinction between the beginninglessness of causal sequences and the beginninglessness of ultimate reality to the exact opposite effect of how Dharmakīrti employs it: instead of denying any connection between the karmic conditions of various sentient beings and ultimate reality, Abhinava claims that all actions are the expression of S'iva's will. In response to an objector who claims that the postulation that all agents are but limited expressions of S'iva's own agency would demolish the distinctions between agents in the conventional world, Abhinava states: "You are right, Sir! For you should know that there is never any effect, which is produced by good or bad [actions]. But those who do not understand it in this way experience [the effects] without [ever] realizing this. For what is called karma is [actually] an impurity (mala), which has ignorance as its source."³⁶ Karma is indeed the reason why individuals experience themselves as existing within a particular world, and this experience is an error. Moreover, the cycles of worlds created in this way are beginningless: "Here [in our system] exists this great creation of S'iva which is replete and inside of which all other [cycles] of creation and resorption take place. It is not proper to say that this is the first creation, for how could something be first etc. in a reality that is without space or time."³⁷ Rather than using this beginninglessness of causal processes to claim that there is no need to provide an ultimately real root cause for diversity, Abhinava uses it to point to the inherent connection between the ultimate and the diverse worlds of conventional experience.

As Ratié points out in a different context, the Pratyabhijñā S'aivas' claim that their position that ultimate reality is the source of both difference and identity in the conventional world allows them to avoid the contradiction into which they accuse Dharmakīrti of falling (Ratié 2014). As we have seen, these S'aivas argue that beginningless karmic imprints that are not inherent in ultimate reality cannot account for the diversity experienced in the conventional world. This critique hinges on the idea that something undifferentiated cannot cause differentiation because it is contradictory for a thing with a purely unitary nature to cause different effects. However, for these S'aivas, it is not contradictory for ultimate consciousness to be the cause of both unity and diversity in the conventional world because ultimate consciousness inherently contains both.

To bring all of this together: time is the expression of the differentiation inherent in ultimate reality by its very nature. This expression manifests as the karmically conditioned experiences of limited subjects, which are themselves formed alongside limited objects via apoha, which is the process of exclusion whereby S'iva carves conventional worlds out of himself. Concepts simply are the expression in time of a particular slice of the inherent differentiation of the ultimate. Perceived, so to speak, from their own sides, both time and ultimate reality are beginningless, but in different ways. The beginninglessness of time is the fact that a first cause of a causal process unfolding in time cannot be identified. The beginninglessness of ultimate reality is the fact that the ultimate is beyond time since time is a limited manifestation of its inherent differentiation. Ultimate consciousness both is the source of time and always exceeds its own temporal expressions. This connection between the ultimate and the conventional in terms of the expression of time is important because it accounts for the way in which it is possible for diversity to manifest at all. Rather than maintaining ultimate reality as pure and utterly removed from the conventional, Abhinava links the two realities in part by claiming that the beginninglessness of the conceptually constructed, causal worlds of saṃsāra relies on the non-dual beginninglessness of the ultimate.

Conclusion

Whether or not Abhinavagupta's own exploration of time is philosophically viable, close attention to the different types of beginninglessness referenced in the Pratyabhijñā debate with Dharmakīrtian Vijñānavāda reveals what is at stake in the S'aiva demand for an explanation of the origin of differentiation. It is not a demand to account for the beginning of a causal process. In their critique of Dharmakīrti's position that a beginningless karmic imprint is sufficient to account for the diversity experienced within the conventional world, the Pratyabhijñā S'aivas do not deny (1) that differentiation among karmic imprints is the immediate cause for differentiated experiences, (2) that the causal processes expressed by these karmic imprints are beginningless, or (3) that it is not possible to identify a root cause for a temporal process from within this process itself. The origin of differentiation does not come from a root cause, such as an original seed producing all subsequent seeds and sprouts. Rather, the origin of diversity must be an expression of the nature of the ultimate stuff that itself constitutes causal processes. Otherwise, causality would be either random or nonexistent, leaving no way to account for the limited realities of various conventional worlds.

In short, according to these S'aivas, rather than rendering the question of the causes of diversity irrelevant, it is precisely because causal processes are beginningless that they must be an expression of the differentiation inherent in ultimate reality itself. Precisely because one cannot identify an initial cause for the string of cause and effect within the conventional world, one must account for the reality of the stuff involved in these causal processes. Since Dharmakīrti relies on beginningless karmic imprints to account for the diversity of the conventional world, he cannot simultaneously maintain that ultimate reality is purely undifferentiated without entering into a contradiction. The incoherence of the question of what exists before time does not mitigate the necessity of accounting for the beginningless origin of time itself, which can only be found in an ultimate reality that both encompasses and moves beyond the conventional world. To the extent that Dharmakīrti's apoha theory relies on beginningless karmic imprints to provide the basic structures of saṃsāra that must be in place before a concept may be formed via exclusion—and this seems to be a significant extent—a defense of his account of concept formation would benefit from addressing this question head-on.