The Nyāyakusumāñjali’s Injection of Revelation into Philosophy: The Role of the First Two Stabakas

Abstract

Scholars universally regard Udayana’s Nyāyakusumāñjali (NK), or Flower Offering of Logic, as one of the great works of classical Indian philosophy, and more specifically, of rational or natural theology. But an important aspect of this masterpiece has not been appreciated by contemporary scholars: Udayana’s many references in the NK to Hinduism’s traditional sacrifices and sacred scriptures are integral to the philosophical case he develops for the existence of a Supreme Being. Or so we contend. We explain our interpretation of the NK through detailed examination of what we take to be the main argument of its first two chapters, an argument that only an extraordinary, omniscient being could have authored the Vedas, the Hindu scriptures. Then we show the importance of this argument for understanding the NK as a whole, including its final chapter, the chapter that has been the focus of most scholars. Though appeal to the Vedas is integral to Udayana’s full argument for the existence of a Supreme Being, his argument is not circular, we maintain; nor was he defending fideism. We believe that Udayana’s approach has relevance for persons of any faith who wish to affirm the centrality of the holy scriptures of their faith to their religious beliefs while recognizing the power of philosophical argument for the existence of a Supreme Being.

1. Introduction

Udayana’s Nyāyakusumāñjali, or Flower Offering of Logic, aims to develop a complex dialectical argument that there exists a Supreme Being (Puruṣottama) whose worship is the means to “heaven and liberation”. Because it mixes reasoning that is clearly philosophical with scriptural references the purpose of which is often obscure, the method is confounding. Many commentators, struggling to understand the relationship between Udayana’s brilliant and clearly philosophical argumentation, on the one hand, and on the other his extensive reference to, and interpretation of, the Vedas, tend to downplay the Vedic references. They may view the reasoning utilizing the Vedas as question-begging (as assuming a Supreme Being’s existence before it is proved), or as merely illustrating the harmony of faith and reason, or as directed toward a very narrow audience, namely, the Mīmāṃsākas.

But these interpretive views are incorrect or incomplete. The key to recognizing an alternative lies in understanding the role of the first two Stabakas or chapters of the Nyāyakusumāñjali. “Stabaka” means cluster or bunch, but here connotes a cluster of arguments. The importance of the first two Stabakas is easily missed in the rush to reach the gorgeous blossoming of the fifth Stabaka—Udayana’s stunning and rightly famous formal arguments, 18 of them, for God’s existence.

Our main conclusions are: The first two Stabakas of the Nyāyakusumāñjali develop a philosophical argument, neglected by contemporary commentators, for the existence of an extraordinary, omniscient being. This argument focuses on the content of the Vedas and the efficacy of the Vedic sacrifices in producing adṛṣṭa, but does not presuppose the Vedas are authoritative and hence avoids the vicious circularity warned against by scriptural Hindu schools. It also constitutes the initial stage of the Nyāyakusumāñjali’s extended cumulative philosophical case that there exists a Supreme Being, the worship of whom is the means to heaven and liberation.

We work toward these conclusions through reflection on the Nyāyakusumāñjali’s first two Stabakas:

Section 2 provides background for our discussion by introducing the Nyāyakusumāñjali (NK), by explaining the challenge of understanding its method of treating scriptural and philosophical arguments in a way that does not impute to Udayana viciously circular reasoning, and by describing how some contemporary commentators on the NK have attempted—unsuccessfully, we maintain—to address this challenge.

Section 3 sets out and elucidates Udayana’s complex dialectical argument in the NK’s first Stabaka, that is, his argument that the Vedic sacrifices are efficacious through adṛṣṭa, understood as trans-empirical merit produced in the soul by performing Vedically based sacrifices.

Section 4 sets out and elucidates key parts of the central argument in the NK’s second Stabaka, that is, his case that, granted the Vedic sacrifices are efficacious through adṛṣṭa, an extraordinary, omniscient being authored the Vedas.

Section 5 concludes by summarizing and extending our argument that the combined reasoning of the first two Stabakas initiates a cumulative case for the existence and worship of a Supreme Being, a case that relies on the content of putative divine revelation but does not appeal to the putative revelation as authoritative, thus avoiding the vicious circularity warned against by the Mīmāṃsākas and Vedāntins.1

Our argumentation is, we think, significant in two ways. First, we offer an interpretation of the first two Stabakas of the NK and their role in the whole of the monograph that highlights aspects of the NK unremarked by contemporary western commentators, including Udayana’s intriguing use of the content of the Vedas in the philosophical case he develops for a Supreme Being, and including his extended reliance on cumulative case argument. Now this may seem a limited significance, one that looks back a millennium to highlight a religious practice—Vedic sacrifice—that will seem curious to most Westerners today, even those of a religious bent, and one we do not ourselves believe in (though many faithful Hindus in India and across the globe still engage in these very ancient sacrifices). But our argumentation has, we hope, a second and broader sort of significance. For although contemporary western theists may not endorse the Vedic content Udayana relied on in the NK, the pattern of his argumentation provides a paradigm they can use to great benefit in developing their own philosophical arguments for God’s existence and for the reliability of a particular revelatory tradition. Udayana’s use of the content of putative revelation as part of a cumulative case points to a general alternative to what we will describe, in coming pages, as the “standard natural theology” that has dominated western 20th- and 21st-century analytic philosophy of religion.

Throughout the NK, Udayana uses “Puruṣottama” (Supreme Being) and “Īśvara” (the Lord) interchangeably, following broader Hindu usage. Most Hindu theists either identified “the Lord” with Viṣṇu or with Śiva (Udayana was a Śaivite). This can be confusing to Western thinkers since those who identified the Lord with Viṣṇu did not deny that Śiva is a divine being of some sort, worthy of reverence, nor did those who identified the Lord with Śiva deny that Viṣṇu is divine. The dispute between them had to do with which of the two is the Lord, the Supreme Being, the “great God”, as they would often say. The dispute was more than merely verbal, each side at times accusing the other of idolatry.

We use both “God” and “the Lord” for “Puruṣottama” and “Īśvara”. Some have objected to calling the Supreme Being whom the Nyāya-Vaiśeṣikas worshipped “God”.2 We understand that concern. Westerners are likely to think of God as creator of the world ex nihilo and also as omnipotent. But no theistic Hindu school of thought (with the possible exception of the Dvaita Vedanta) believed that creation from nothing was possible, and the Nyāya-Vaiśeṣikas held that Īśvara could not, in some sense of “could not”, ignore the karmic deserts of finite souls, and so, one might argue, was not omnipotent. However, our use of “God” and “the Lord” for “Puruṣottama” and “Īśvara” follows the approach of most contemporary Western scholars. Our thought is that, from a religious point of view, to be God is to be a person worthy of worship who possesses to the maximum degree possible every attribute that is absolutely speaking better to possess than not. Under that definition, the Īśvara of the Nyāya-Vaiśeṣikas seems to count as God.

2. Background

2.1. The Nyāyakusumāñjali

Little is known about Udayana’s life, as unfortunately is generally the case with classical Indian philosophers. Even his dates are the subject of some dispute, though most scholars now hold he flourished in the province of Mithila between the eleventh and twelfth centuries. But it is beyond dispute is that he was one of the greatest thinkers of the Nyāya-Vaiśeṣika school, perhaps the greatest. The Nyāya and Vaiśeṣika were originally sister schools, probably stemming from a common root (Potter 1977, pp. 12–13 and 21–23; Rostalska and Bilimoria 2025, pp. 122–23). Though their disagreements were always slight, they had different foundational texts, the Nyāya Sutras for the Nyāya and the Vaiśeṣika Sutras for the Vaiśeṣika.

Udayana composed three commentarial and four non-commentarial works.3 In his last two non-commentarial works, the Ātmatattvaviveka, or Concerning the True Nature of the Self, and the Nyāyakusumāñjali, Udayana completed his novel defense of central Nyāya-Vaiśeṣika doctrines. The first, in spite of its title, is devoted to defending Nyāya-Vaiśeṣika substantivalism in general, especially against Buddhist attacks (Potter 1977, p. 7; Chakrabarti 1999, pp. 219–76). In it Udayana defends the existence of atomic material substances, composite material substances, finite souls and finally, the existence of the infinite soul—Īśvara, the Lord.4 The second, the NK, is focused on arguing for the existence of Īśvara. In spite of obvious similarities between these works, their methods of argument differ in subtle ways, ways that have not been fully understood or appreciated in the secondary literature written on them in English, but that we will comment on in this paper.

The NK consists, technically, of a set of verses: seventy-three, to be exact. Nine of these—the opening passage, the concluding prayer of each of the five Stabakas, and the final passage—are quite moving: they breathe with genuine piety. The majority of the other verses tersely summarize the dialectical arguments Udayana intends to give in answer to certain questions and objections. He wrote his own commentary on these verses, called the Prakarana, and without that, the purport of the technical verses of the NK would be difficult, if not impossible, to ascertain. The NK plus Udayana’s auto-commentary is the Nyāyakusumāñjali-Prakarana. “Prakarana” here connotes a unified interpretation of a text. Understandably, when most scholars refer to the NK, they mean to refer both to the verses and Udayana’s auto-commentary, unless they specify otherwise.

Classical Indian philosophy was as dialectical as medieval Western philosophy. Philosophers from schools that argued for substantive metaphysical and religious positions embraced a dialectical method; even the Advaita Vedānta, in spite of the school’s teaching that the ultimate reality is beyond the realm of conceptual categories, gave philosophical arguments for certain metaphysical positions (Forsthoefel 2018, pp. 236–38). The method was also employed by philosophers of a skeptical bent, such as Jayarāśi (Balcerowicz 2017, esp. pp. 406–11), and by thinkers from the Madhyamaka school of Buddhist thought, a school that denied the principle of the excluded middle (Ganeri 2001, pp. 47–51). The majority of classical Indian schools accepted a pramāṇa-based epistemology—i.e., a means-of-knowledge-based epistemology.5 Philosophers in the pramāṇa tradition proceeded by asking questions of the form: “Is it the case that P, or ~P?” The masters of Indian philosophy would immediately disambiguate where necessary, stating precisely in what sense the question was being asked. Then the favored answer would be supported through an argument, or several of them. Invariably, objections would be raised and answered. Frequently, “counterbalancing arguments” for the opposite answer would be given and objections raised to those. These objections were ultimately sustained.

Though a dialectical method has its advantages, it is often hard to see the forest for the trees—indeed, the works of great masters of the approach are vast forests one can easily get lost in. The NK is a dense dialectical labyrinth, partly due to the number of views that answer to the range of questions Udayana takes up, and perhaps partly because Udayana pursued some issues further than necessary, given the NK’s aims: his desire at certain points to defend the Nyāya school’s solutions to various abstruse metaphysical and epistemological questions increases the odds one will get lost in the labyrinth.

Fortunately, Udayana gives the attentive reader navigational guides. An initial signpost comes in the second verse of Stabaka I. In it, Udayana articulates the NK’s central topic, saying that “The Supreme Being, whose worship the wise consider to be the way to heaven and liberation, is being considered here.”6 This early topic statement, together with its subsequent development, suggests the NK argues for a two-part central thesis:

CT:

There exists a Supreme Being, the worship of whom is the means to heaven and liberation.

Many ignore the fact that it is not the existence of the Supreme Being taken in abstraction that is considered in the NK, but rather the existence of a being “whose worship the wise consider to be the way to heaven and liberation”. The two parts of this central thesis—first, that there exists a Supreme Being, and second, that worship of the Supreme Being is the means to heaven and liberation—form a compound analogous to that of “conjoined twins” (formerly called Siamese twins). There is a symbiotic relationship between CT’s conjuncts: the arguments for the two propositions entwine; each proposition helps support the other. The strength of the symbiosis enables Udayana to build his cumulative case argument for God’s existence, a case that integrates reason and revelation.

Almost immediately after articulating the focus of the NK, Udayana goes on to say there are five views that oppose the position expressed by CT. Each Stabaka takes up one of the opposing views:

• There are no means to another world.
• Even if there are means to another world, we could know of them without a revelation by any omniscient, super-human being.
• There are conclusive reasons for thinking that no Supreme Being exists.
• Even if an unembodied intelligent being exists, he could not be a source of knowledge for us since, lacking a body, his cognitions could never be warranted/justified.
• There are no reasons for thinking that a Supreme Being exists.7

2.2. The Threat of Vicious Circularity in the NK

Some classical Hindu philosophical schools, especially the Mīmāṃsā and the various Vedānta schools, saw themselves as primarily engaged in what contemporary thinkers term “philosophical theology”. They used philosophical concepts and methods to interpret and defend the Vedas, the Hindu “Bible”.8 In the process, they defended some of the basic ontological and epistemological doctrines of their schools. But when it came to what might be thought of as the deepest questions concerning the fundamental nature of reality, they held that the Vedas alone should guide us. For example, theistic Vedānta schools denied the probative power, and even the soundness, of arguments for God’s existence that do not appeal to the authority of the Vedas in their premises (Raghavachar 2016, pp. 72–73; Lipner 1986, pp. 3–4).

The Nyāya-Vaiśeṣikas had a contrasting approach. They thought we could know much more about God’s existence and nature through reason alone than did the Vedāntins, though they would at times supplement reason-based arguments for various positions by appeal to the Vedas. For example: Gautama, in his Nyāya Sutras, and the early commentators on it, first argue philosophically that human beings have substantival selves, selves constituted by an eternal, immaterial soul; then they add a karma-based argument for the existence of such a self that clearly relies, ultimately, on the authority of the Vedas (Dasti and Phillips 2017, pp. 84–85). It is partly because the Nyāya-Vaiśeṣikas relied so little on the Vedas with respect to fundamental metaphysical and epistemological doctrines that they were held by more scriptural Hindu schools to be only semi-orthodox. Indeed, the older Nyāya-Vaiśeṣikas relied primarily on the authority of the Vedas only with respect to certain moral and religious duties incumbent on a believing Hindu (Dasti and Phillips 2017, pp. 157–58).

Though the Nyāya-Vaiśeṣikas favored a philosophical approach in their works, they endorsed the following propositions:

• We know that God exists because we know the Vedas are authoritative and their content is true.
• We know the Vedas are authoritative and their content is true because we know the Vedas were revealed or authored by God.

Adherence to the second of these propositions seems to have developed gradually. The Nyāya Sutras trace their authority to the rishis who “heard” them, saying nothing about their authorship by God. We surmise that the later Nyāya came to see that the rishis could not, by yogic meditation alone, come to know of transcendental matters without divine inspiration, and hence that God should be taken to be the true author of the Vedas, the rishis being but scribes.

The Mīmāṃsākas and the Vedāntins held that these two theses constitute, or at least threaten, a vicious circle (Chemparathy 2018, p. 7). The theistic Vedāntins, who agreed with the Nyāya-Vaiśeṣikas in maintaining the first thesis, attempted to avoid a circle by teaching, along with the atheistic (or non-theistic) Mīmāṃsākas (Bilimoria 2024, pp. 426–28), that the Vedas are authorless and eternal (though they held the Vedas depend for their very being on God), in this way upholding a kind of “biblicism” far more radical than that taught by some fideistic Christian and Muslim theologians (Raghavachar 2016, pp. 32–33; Pereira 1976, pp. 84–85). The more scriptural Hindu theists further argued that the Nyāya-Vaiśeṣikas could avoid reasoning in a circle only if they first showed, using arguments making no appeal to the authority of revelation, that there exists an omniscient, omnipotent, perfectly good spirit who created and sustains the world, and then showed that this spirit would want to reveal things to human beings for their moral and spiritual well-being, and finally, that the Vedas fit reasonable criteria for having been authored by this spirit (Raghavachar 2016, pp. 32–33). In this approach, the content of putative revelation is not identified as important evidence that needs to be considered early-on in an investigation of the existence of an omniscient, omnipotent, perfectly good spirit, a creator and sustainer of the world.

A spirit with these divine-making properties might be called a “big-G God”. We understand a “big-G God” as a being with the full range of divine-making properties, whatever those properties are taken to be. We take it that the scriptural Hindu theists, and the Nyāya-Vaiśeṣikas, and believers in the Judaeo-Christian tradition all accept the existence of a big-G God.

The general method which, according to the Mīmāṃsākas and Vedāntins, the Nyāya-Vaiśeṣikas would have to employ in order to avoid a circle is so prevalent among contemporary analytic philosophers who think of themselves as doing natural theology that we might characterize the method of standard natural theology (SNT) as an approach that follows the investigative ordering protocol described above, an approach that does not identify the content of putative divine revelation as important evidence for investigating the existence of a big-G God, but rather saves it for consideration until after the existence of such a being has been established.9

Udayana himself followed the method of SNT in the Ātmatattvaviveka (Dravid 1995, pp. 380–429; Chemparathy 2018, pp. 24, 87). Since Udayana had used SNT in the Ātmatattvaviveka, and since his doing so accords with illustrious works of earlier thinkers of his school, such as Śrīdhara and Jayanta Bhaṭṭa (Jhā 1982, pp. 116–28, 455–65; J. V. Bhattacharya 1978, pp. 406–26, 484–22), it makes sense to think he would follow the method in the NK as well. And, indeed, all the scholars we know of who have attended to the NK hold that it does employ SNT (though they don’t use that label). But a careful examination of the whole of the NK presents deep problems for this interpretation, since most of the arguments in the NK for any sort of God-like being make reference to the Vedas in one way or another, and it is not evident that the references are merely decorative or non-essential.

In light of this, scholars who have carefully studied the NK and who believe Udayana used SNT in writing it have proposed various ways of understanding its references to scripture. N. S. Dravid, for example, holds that Udayana simply begs the question against atheists in most of the arguments of the NK for the existence of God that bring scripture into the picture: the charge of circularity is accepted (Dravid 1996, pp. 419, 432, 488). But granted Udayana’s acknowledged skill as a logician, one should be reticent to accuse him of such a blunder, particularly since begging the question (or something very like it) had long been recognized by classical Indian logicians as a fallacy. Other scholars, perhaps most notably Sinha, suggest that Udayana intended the scriptural arguments to show the harmony of faith with reason.10 Though this proposal works for some of the theistic arguments of the NK, it will not work for many of them. Chemparathy takes a somewhat different approach, suggesting that in some cases Udayana was using scripture in the way Sinha suggests, but that he often was arguing, against the Mīmāṃsākas and other atheistic Hindus, that contrary to what they thought, the Vedas do teach that a big-G God exists: the injunctive passages of the Vedas, for instance—the only ones the Mīmāṃsākas thought important—don’t make sense unless we suppose them to be commands of a good God (Chemparathy 2018, pp. 79–80, 185).

We think that Chemparathy’s approach is correct so far as it goes, but does not appreciate the fact that Udayana’s extended argumentation in the NK for God’s existence uses a subtler approach than SNT. This subtler approach integrates reason and revelation in part through rejecting the key proviso of SNT that the content of putative revelation is to be saved for consideration until after it’s been established that a big-G God exists. The subtler approach, which rejects the key proviso of SNT without falling into vicious circularity, is signaled in the oft-neglected argumentation of the first two Stabakas, which relies on the content of putative revelation as evidence for God’s existence.11

3. The NK’s First Stabaka

3.1. AEVS: The Argument for the Efficacy of the Vedic Sacrifices

The objection Udayana takes to motivate and inform Stabaka I (“There are no means to another world”) was put forward by the Cārvāka school of ancient Indian philosophy. The Cārvākas were very like the ancient Greek atomists. They held that the ultimate building blocks of reality are atoms of earth, air, fire, and water. The human person is an aggregate made of these. Consciousness is a supervening property of the sense organs as a whole, or the inner sense organ (something analogous to the brain), in the way that the intoxicating power of fermented barley is a supervening quality of beer. The world comes to be without design since there is no intelligent agent who made it on purpose for a reason. When you are dead, you are dead; religious doctrines were invented by charlatans who wished to fool people in order to gain wealth and power. The Vedas are internally inconsistent and contradict what we know by perception. In all these ways, the Cārvāka teaching was similar to that of the Greek atomists. But there is an aspect of the Cārvāka doctrine that cannot be found in Greek atomism: they raised proto-Humean worries about induction and causality. They used these worries, for example, to undercut the sorts of design arguments the Nyāya put forward to defend the existence of an unembodied intelligent agent who made the earth, the planets, trees, etc. And they used them as well to argue that there are no means to another world (or, at least, that we could never have good reason to believe there are such means).

Strains of a basically materialistic world view can be found even in the Vedas and the Upanishads, and materialist philosophical schools arose in India during the 5th/6th centuries BCE, a time of great philosophical and religious upheaval. The Cārvāka seem to have been the last of these schools—at any rate, they were the only school of materialism that later philosophers from various other schools bothered to discuss. By the time Udayana wrote, the Cārvāka had long ceased to have adherents, but Hindu, Jain, and Buddhist philosophers still felt the need to refute various Cārvāka doctrines and arguments.12

In Stabaka I, Udayana set out to defend the view that there are means to another world, indeed to heaven and liberation. In that connection, he sought to defend doctrines concerning karma, rebirth, and liberation that had become central to Hindu, Buddhist, and Jain schools of thought. (The karma theory was not part of the ancient Vedic religion, and only hints of it can be found in certain later Upanishads. Furthermore, the oldest teachers of the Mīmāṃsā school, arguably the most orthodox of orthodox Indian philosophical schools, did not teach the doctrines of karma, reincarnation, the periodic creation and destruction of the world, or mokṣa (Halbfass 1980, pp. 271–72).)

Udayana aimed to defend the view that there are means to another world in a way that accorded with the beliefs of his school, beliefs that were not held by all Hindu schools. He did so by defending the efficacy of the Vedic sacrifices, whether the sacrifices were aimed at material rewards for the sacrificer, or at cleansing from sin, or at rebirth in heaven. And he defends the efficacy of the sacrifices by appealing to, and defending belief in, the existence of what the Nyāya-Vaiśeṣikas called adṛṣṭa. “Adṛṣṭa” literally means “the unseen”. The roots of this teaching go back to the sacrifices made to the devas (“shining ones”) of the Ṛg-veda. The devas were hypostatizations of natural forces and abstract concepts, and were held to rule by Ṛta, a cosmic moral force. It was believed that by sacrificing to them, some of the power they had expended in their cosmic sacrifice, a sacrifice that produced the world, would be restored to them (Bilimoria 2021, pp. 187–88; Boyce 2001, pp. 3–12). One of the most subtle of the later speculations this gave rise to is contained in Śabara’s commentary on Jaimini’s Mīmāṁsā Sūtra: the sacrifice, it is suggested, produces apūrva, a power that will ensure the sacrificer will enjoy the fruit of his action. Later Mīmāṃsākas became interested in the ontology of this power, e.g., in whether it resides primarily in sacred objects, or the soul of the sacrificer, or the sacrifice itself, or some combination; in the course of their discussion, they fashioned a sophisticated causal theory, one that took causal power to be an irreducible ontological category (Radhakrishnan 1931, pp. 376–77, 421–22; Pereira 1976, pp. 89–95; Chari 1988, pp. 365–69; Halbfass 1980, pp. 274–77; Bilimoria 2021, pp. 183–88). The majority of the older Nyāya-Vaiśeṣikas had long held that the causal powers of things are reducible to the essential or contingent qualities they have and, in line with this, they attempted to undercut the general causal theory that grounded the Mīmāṃsā doctrine of apūrva (Matilal 1997, pp. 289–94). Udayana weaves a complex dialectic in Stabaka I of the NK in defense of the leaner Nyāya-Vaiśeṣika theory of causal powers, in this way giving support for the Nyāya-Vaiśeṣika view that the sacrifices are efficacious through adṛṣṭa, an imperceptible (save to God) quality of the soul, not apūrva.

Udayana’s precise understanding of adṛṣṭa was bound up with the sophisticated metaphysics he inherited, defended, and developed with matchless skill over the course of his writings. He took adṛṣṭa to be the supersensible merit or demerit produced in a person’s soul by their doing something morally or religiously good, or by their doing something positively wrong or failing to do something obligatory; it is a quality that brings with it a directedness toward being rewarded, or punished, by a trans-empirical moral power that will ensure that souls will be blessed in accordance with their merits and penalized in accordance with their demerits, whether in this life or in a future life. The Nyāya, not unlike the Buddhists, took karmic suffering to be educative; it was not crudely retributive (Chemparathy 2018, p. 160). As we articulate and defend Udayana’s argument in Stabaka I for the existence of adṛṣṭa we will have more to say about its nature.

How does that argument go? It is entwined with Udayana’s argument for the efficacy of the sacrifices: once again, we have conjoined theses, with interwoven supporting arguments. The conjoined theses are that The Vedic sacrifices are efficacious, and that If the Vedic sacrifices are efficacious, it is adṛṣṭa that makes them so. In combination: The Vedic sacrifices are efficacious through adṛṣṭa. Abstracting from the dialectical, piecemeal method Udayana employs, and leaving to one side many fascinating byways, we may formalize the argument (call it AEVS, for Argument for the Efficacy of the Vedic Sacrifices) as follows:

• Many people, for a long time, have performed Vedic sacrifices to effect various results, above all to be reborn in heaven and to wash away sin.
• If (1), then the Vedic sacrifices are efficacious unless there is a defeater for belief in the efficacy of the sacrifices.
• There is no defeater for belief in the efficacy of the sacrifices.
  • Specific Cārvāka criticisms of the efficacy of the sacrifices are unsuccessful.
  • The general Cārvāka criticism that attempts to undercut the efficacy of the sacrifices consists in their defense of materialism; but their defense of materialism is untenable.
• If the Vedic sacrifices are efficacious, they are so through adṛṣṭa.
• [So,] The Vedic sacrifices are efficacious through adṛṣṭa (Dravid 1996, pp. 26–27; cf. Sinha 1999, pp. 31–33; Joshi 2002, pp. 45–47; Cowell 1864, pp. 9–10).

3.2. Premises (1) and (2) of AEVS

The first premise of AEVS would have been taken as obviously true by Udayana’s audience: they knew first-hand the long-established practice of Vedic sacrifice for the sake of acquiring merit and getting rid of demerit. The Hindus saw the sacrifices in general, which were to be performed every day, as a way of worshipping God (or the gods) and effecting contact with the divine. In all of this, correct moral purpose on the part of the sacrificer, purity of heart, was key. For Udayana, to say that the sacrifices are efficacious is to say they produce the effect aimed at: placating God, bringing rain, generating a son, atoning for sin, acquiring spiritual goods, and so forth. The most important petitionary sacrifices—sacrifices for a purpose other than praising or thanking God—were directed to the afterlife: people aimed to be reborn in a higher form. Ultimately, the aim was spiritual enlightenment (mokṣa, that is, release).

What reason is there for accepting premise (2)? In seeking evidence for the efficacy of the sacrifices, Udayana places great emphasis on the beliefs and motives of the multitude of people. He does so in part because the most important sacrifices have their effects—in the person making the sacrifice—after death, so one can’t prove the efficacy of these sacrifices by observing that the sacrifices have produced their effects, by saying, for instance, something akin to: “Look: the harvest was good”, or “Look: he had a son”. And more generally, Udayana held that one couldn’t empirically prove that the sacrifices have certain effects. But he also thought, like Reid, that universal belief is a mark of truth. If one questions a universally held belief, one is questioning a natural human belief. And if one questions such a belief, there is no stopping point: it leads to total skepticism.

3.3. Premise (3-a) of AEVS

But the Cārvāka had multifaceted lines of reasoning specifically targeting the efficacy of the long-standing sacrificial practices. One aspect of their argument involved criticisms of the motivations and beliefs of the sacrificers:

The Cārvāka suggested that people may always perform sacrifices to gain a visible reward, rather than for merit in an afterlife.

–

Udayana responded that the most important Vedic sacrifices don’t even promise a visible reward.

Maybe, the Cārvāka said, people perform sacrifices to gain fame or respect, rather than because they believe the sacrifices are efficacious in the way the Vedas assert.

–

But, Udayana countered, this would not even be possible unless most people believed in the sacrifices, and besides, the most important sacrifices are done in private and the most pious of people don’t care anything for fame or respect or riches.

Perhaps, the Cārvāka proposed, people perform sacrifices because it gives them pleasure.

–

But in that case, even non-believers would perform the rituals, Udayana maintains.

Or perhaps people sacrifice out of convention.

–

But Udayana counters: The majority of people and the very oldest texts hold that all sacrifices have some good result, at the very least the washing away of sin.

Or maybe people were just duped by their elders.

–

But the elders themselves perform the sacrifices, thus on this account they must have been duped by their elders, and so on, and so on. Some source of the duping must be found, since erroneous cognitions arise from some fault, and if one holds otherwise then universal skepticism follows.

Maybe, though, there was an original duper or dupers, way back in the past, who enjoyed cheating people, or wanted fame, as a cult leader does.

–

But how could pleasure in duping others make up for the trouble the original dupers would have to go to, to convince others of the efficacy of the sacrifices? Besides, what we know of the seers of old suggests that they were selfless and did not care for riches or fame (cf. Dravid 1996, pp. 25–27; Sinha 1999, pp. 31–33; Joshi 2002, pp. 45–47; Cowell 1864, p. 9).

Missing from Udayana’s litany of Cārvāka concerns is a particular criticism we might label “the contemporary objection”. Why not simply say that the Vedic religion, which was built up gradually over time, was based on a non-culpably erroneous understanding of natural phenomena? The majority of people today do not believe in the Vedic sacrifices, and that may seem to provide traction for the contemporary objection. We do well to remember, though, that most, and maybe all religions, did at least initially involve sacrifices, and most people throughout history have been religious. These are deep and underappreciated phenomena. We believe that the universality of the religious impulse is in the background of Udayana’s interesting remark that false religion requires true religion, and of the more general doctrine of the Nyāya that falsity is grounded in truth and that illusions and false beliefs are mis-combinations of things that are real, or beliefs that are true. There’s a basis of truth underlying all falsity and false belief. Applied to the religious impulse, we could say that to suppose that all religious beliefs are false is to suppose that the natural cognitive faculties of human beings give rise to beliefs that are thoroughly delusory; but that might be thought to give rise to doubts about the reliability of the cognitive powers of human beings, leading to global skepticism.

3.4. Premise (3-b) of AEVS

While the Cārvāka doctrine of materialism does not specifically target the efficacy of the sacrifices, it entails that the sacrifices aren’t efficacious. Udayana argues in Stabaka I that the Cārvākas’ defense of their materialistic worldview is untenable. And if it is untenable, then the view that the sacrifices are efficacious becomes considerably more plausible. In and of itself, Udayana’s argumentation here only establishes a necessary condition for the efficacy of the sacrifices. But that’s important. Consider how astronomers looking for life on other planets get excited if they discover that there is—or was—water on other planets. Water doesn’t prove life; but it’s a necessary condition.

The subtlety and sophistication of Udayana’s reasoning in attacking materialism is almost impossible to overstate. To hint at its depth, we’ll summarize part of his attack on a materialistic account of the origin of the universe, and describe part of his attack on the view that consciousness is an emergent property of composite material substances.

One aspect of Udayana’s attack on materialism consists in his explanation of why the world could not have come into existence accidentally.13 He reasons that if the world were to come into existence accidentally, it would have to do so in one of the following ways, and, he argues, it couldn’t have come into existence in any of these ways.

• The world comes to be “accidentally” in the following sense: there does not exist anything, in any sense, from which the world comes.
• The world comes to be “accidentally” in the following sense: it has causal antecedents, but they don’t produce the world. (This option seems to present something like Hume’s constant conjunction theory of causality: it seems to entail that all there is to supposedly necessary causal connections is the constant conjunction of events of certain types. Udayana takes it that there’s more to causality than this and later drives the point home by saying that even if every time we’ve seen smoke, a donkey is near, the donkey is not causally relevant to the smoke, at least not immediately. In India there was a constant conjunction of smoke with donkeys, because washer-persons used donkeys to pull firewood for the washing, and the fire created smoke.)
• The world comes to be “accidentally” in the following sense: it causes itself. (This is an option for which the atheistic philosopher Quentin Smith argued (Smith 1999).)
• The world comes to be “accidentally” in the following sense: it is caused by “nothing”, i.e., by something absolutely unreal, such as a round square.
• The world comes to be “accidentally” in the following sense: it is its nature to come into existence at a particular point in time. (According to this view, the world explains its own existence, but the world doesn’t, strictly speaking, cause itself. One might think here of Leibniz’s idea that God’s essence explains his existence but doesn’t cause it.)

A complex dialectic supports Udayana’s rejection of all these possibilities.

Consider, for instance, his criticism of the first option (that there is not anything, in any sense, from which the world comes). To exist, a thing must at least be absolutely, i.e., metaphysically, possible. But either its possibility is also sufficient for its existence, or not. If the metaphysical possibility of something existing is sufficient for it to exist, then everything that can exist, necessarily will exist. Hence there will be no contingent beings. And hence there will be no non-eternal beings. But there are non-eternal beings. So, the metaphysical possibility of some thing’s existing is not sufficient reason for it to exist, and if that is the case, it has need of some cause to produce it (Dravid 1996, p. 10; cf. Sinha 1999, pp. 14–15; Joshi 2002, p. 35).

Or consider some of his argumentation against the fifth option, that it is the nature of the world to come into existence at a particular point in time. This is the possibility he spends the most time on. He imagines how the Cārvāka might defend the claim. The Nyāya hold that a clay pot is caused by clay, staff, wheel, and the potter, and that when the pot comes to be, it comes to be in clay “particles” of a certain sort—not, say, in threads, or filaments of cloth (the Nyāya held that clay is earth and so the ultimate constitutive particles of a clay pot would be “earth atoms”). If one asks why that is so, the Nyāya can only answer by appealing to the nature of the pot. But then, the Cārvāka insistently ask, why not say that the pot comes to be at a certain time by its own nature without appealing to any external cause? If its nature, by itself, determines where it comes to be (i.e., in clay particles) it could also determine when it does, it would seem. But, Udayana says, “This will not do!” Because if a contingent thing’s coming-to-be has nothing to limit or determine its time of origination (or destruction), or if what limits it is random and totally unpredictable (i.e., is not regulated by universals), then it could always have existed. But if it could always have existed, it could have been a non-contingent thing—and that is against its very nature. Furthermore, since pots are made from lumps of clay, and in clay particles, particles which are their inherences causes, the fact that they can come to be only where such particles exist is perfectly intelligible based on their very essence. But clay particles don’t, and can’t, turn themselves into pots—that requires efficient causes of some sort, e.g., a potter (Dravid 1996, pp. 10–11; cf. Sinha 1999, pp. 16–17; Joshi 2002, pp. 35–38). Here, and in other places, Udayana is clearly making use of modal concepts, even though the Hindu philosophical tradition in which he was trained did not really supply him with a modal vocabulary.

Another materialist position Udayana attacks is the view that consciousness is an emergent property of composite material substances, specifically of living bodies (human and animal). The Indian materialists held we have strong support for this since whatever we have good reason to believe to be conscious is a living body, and whatever we have good reason to believe to be not conscious is not a living body. But if selves are living bodies, there is no adṛṣṭa, since adṛṣṭa has effects on a person after death and so requires that the self is an immaterial, simple substance.

The Nyāya-Vaiśeṣikas, over time, elaborated various arguments against this kind of physicalism, and Udayana developed a novel argument for the existence of immaterial selves in the last part of his Ātmatattvaviveka. In the NK he briefly defends one of the mereologically based Nyāya arguments against physicalism. This is based on the axiom that “no one remembers what another experienced” (from the inside, that is) and that people do remember things that happened to them days, months, and even years before. Such occurrent memories, however, depend on impressions left by past experiences. Without these impressions, no actual memories of the form “I remember doing x” or “I remember seeing y” would be possible. Nor can these impressions be passed from one person to another, since they are property particulars and property particulars can’t go flying from one substance to another. However, a human body does not remain the same from year to year (or even from day to day), since human bodies are continually changing with respect to the micro-particles14 that make them up, i.e., they lose some and gain others, even if the micro-particles lost and gained are all (typically) of the same sort. Hence our bodies are not numerically the same bodies as they were a year (or, perhaps, even an hour) ago (Dravid 1996, pp. 76–79; Sinha 1999, pp. 54–55; Joshi 2002, pp. 66–68; Cowell 1864, pp. 18–19).

This argument assumes mereological essentialism. Udayana does not defend mereological essentialism in the NK (though he does in other places, and it had long been developed and rigorously defended by the Nyāya-Vaiśeṣikas). Furthermore, the Nyāya-Vaiśeṣikas had a number of arguments for the immateriality and immortality of the soul. Udayana defended many of these in the Ātmatattvaviveka.15

We can imagine that at this point that many contemporary scholars will be unimpressed. Even though the falsity of Cārvāka materialism is a necessary condition for the efficacy of the sacrifices, Udayana’s argumentation hardly shows that materialism of every sort is false. And even if materialism is false, that does not prove the Vedic sacrifices are efficacious, even coupled with Udayana’s argument based on his understanding that people had, for generations, engaged in the sacrifices. There surely were many situations where the sacrifices were performed, but the desired results did not appear.

In response, we point to three considerations.

First, Udayana, like all Nyāya-Vaiśeṣikas, was a fallibilist. The Nyāya held that the only truths we can be certain of concern the existence of mental states we are currently experiencing. In this they agreed with Descartes; but they did not at all agree with his demand for certainty. Udayana would freely have acknowledged that no irrefutable proof of the efficacy of the sacrifices can be given—but then, no irrefutable proof of the existence of an external world can be, either. As he notes in one part of his criticism of Mīmāṃsā internalism in Stabaka II, not all doubts are reasonable. If one has drunk water from a well for many years and is perfectly healthy, the mere possibility that the well contains an undetected poison is not a good reason to stop drinking from the well—especially if there are no other water sources it would be practical to draw from. So too, if one has practiced the religious rituals of one’s tradition for years and found them an aid to the cultivation of moral virtue and to one’s powers of attention, and if generations in one’s community have done the same, one has the epistemic right to believe in the efficacy of such rituals and to continue to practice them, unless a powerful defeater is presented (Dravid 1996, pp. 114–28; Sinha 1999, pp. 75–79; Joshi 2002, pp. 89–96. See also Phillips 2012, pp. 17–32). Udayana devoted much of his philosophical genius to arguing there are no such defeaters. But there are limits to the force of an appeal to Udayana’s fallibilism. There is no getting around the fact that although some doubts are unreasonable, belief in the existence of an external world is hardly on the same footing as belief in the efficacy of the sacrifices.

A second consideration, which helps address concern about the strength of Udayana’s case for the efficacy of the sacrifices, is that his case, and the arguments that build on it, constitute only one strand in his cumulative case in the NK for God’s existence. We are focusing on this strand because we think it has been neglected, and because we think taking it into account will cast a new light on the whole of the NK, showing that Udayana is not employing the standard approach to natural theology (SNT) to defend the rationality of believing that a purportedly revealed text was, in fact, revealed. We believe that the pattern of Udayana’s argumentation provides a paradigm that contemporary philosophers of religion are well-advised to use in developing their own philosophical arguments for God’s existence and for the reliability of a revelatory tradition.

And third, though the NK covers a lot of ground, it was not meant to be a complete defense of Nyāya-Vaiśeṣika philosophy. In response to a challenge from an unconvinced materialist, Udayana presumably would have referred to the corpus of his works, including his commentarial works. In these, one will find a much more thorough attack on all forms of materialism, as well as novel defenses of a sophisticated form of mind–body dualism.

3.5. Premise (4) of AEVS

The first three premises of Udayana’s argument that the Vedic sacrifices are efficacious through adṛṣṭa get us to the sub-conclusion that the Vedic sacrifices are efficacious. We turn, now, to the conditional claim of the fourth premise of AEVS: If the Vedic sacrifices are efficacious, they are so through adṛṣṭa.

Why think there is any such thing as adṛṣṭa? The case for its existence weaves through the intricate dialectic of Stabaka I. Udayana, quite early in the NK, tersely summarizes that dialectic:

From dependency, from eternity, from diversity, from universal practice, and from the apportionment of pleasure and pain to each individual soul, mundane happiness and suffering implies the existence of a supernatural cause (adṛṣṭa).

(Cowell 1864, pp. 4–5)

It is clear that Udayana did not mean to suggest that any one of the arguments he briefly refers to in this dense passage is an independent proof of the existence of adṛṣṭa. As Hari Dasa has implicitly noted, Udayana indicates in this passage that he intends to build a cumulative case argument for adṛṣṭa.

In Hari Dasa’s excellent summary, the phrase “from dependency” was meant to alert us to the complex dialectic for the view that everything that comes to be, has causes, among them some efficient cause or causes. Though Udayana develops this into an argument that the world (the present cosmos) has certain trans-empirical causes, nothing he says in this section of Stabaka I shows adṛṣṭa is among them. The phrase “from eternity” refers to his response to an objection to the view that everything that comes to be, has causes. “From diversity” summarizes his reasons for rejecting certain doctrines of the Adviata and Sāṃkhya Hindu schools. If they are right, then even if adṛṣṭa exists, it is not what the Nyāya-Vaiśeṣika hold it is. Furthermore, their views rule out the existence of a big-G God. “From universal practice” forms the heart of Stabaka I and contains Udayana’s defense of the existence of adṛṣṭa based on the ancient sacrificial practices of the Hindus, practices which Udayana argues make no sense unless adṛṣṭa, as the Nyāya-Vaiśeṣikas conceived of it, exists. Finally, with the phrase “from the apportionment of pleasure and pain” Udayana indicates he will criticize the Mīmāṃsā and the Sāṃkhya views of adṛṣṭa, views that were opposed to the theism of the NK, as well as criticize the Cārvāka doctrine that the self is either the body, or some proper part of the body, such as the brain (or something analogous to it).

Udayana does not provide an argument for adṛṣṭa independent of the argument establishing the efficacy of the sacrifices. Rather, he tries to show that adṛṣṭa is a necessary link between performing the sacrifices and being rewarded by a transcendent moral power. He gives us a conjoined argument that establishes the existence of adṛṣṭa at the same time it establishes the efficacy of the sacrifices.

Here is a way to think about the matter. One could hold that if a person performs a particular sacrifice with purity of heart, then a transcendent moral power will reward the person without need for intervening merit. The Buddhists took it that moral responsibility could be accounted for without an enduring self, in virtue of certain close causal relationships between prior and later mental states in the same stream. Although toward the end of Stabaka I of the NK Udayana does present a novel argument against the Buddhist doctrine of momentariness (Dravid 1996, pp. 82–98; Sinha 1999, pp. 55–66; Joshi 2002, pp. 68–72), his extended case against the Buddhist “no-self” doctrine is to be found in his Ātmatattvaviveka, in which he defends the existence of an enduring self against several Buddhist lines of reasoning. At any rate, Udayana held that the Buddhist doctrine of momentariness, according to which there is no enduring self, does not work: we need an enduring self to link the sacrifice to the reward.

But do we need anything else, beyond an enduring self? Udayana thought we do—he thought that unless there is some supersensible merit produced in the soul of a person that links the sacrifice he performs to the reward, things would be no different than if he had not performed the sacrifice—at least with respect to his being deserving of the reward.

Hari Dasa gives a helpful analogy to further explain Udayana’s argument. In order to remember something which one did, or experienced, in the past, certain traces of what one did or experienced must be produced in one’s brain. Or, more literally, in one’s “inner sense organ” (again, which we take to be analogous to the brain). The inner sense organ was conceived of as a material, non-conscious instrument of the soul, located in the head. More fully, the analogy is as follows: Just as doing or experiencing certain things produces in our “brains” traces or impressions of what we did or experienced, traces that are necessary for us to remember what we did or experienced, so performing the sacrifices produces the meritorious qualities in our souls, qualities that are necessary for us to deserve future rewards and, hence, for the transcendent moral power (whatever that is) to justly reward us (Cowell 1864, p. 10).

Finally, we note that Udayana wanted to defend the reality of adṛṣṭa in a way that would respond to Mīmāṃsā and Sāṃkhya doctrines of the nature of the efficacy of the sacrifices. The Sāṃkhya thought the sacrifices are efficacious by producing an unseen quality that resides in the inner sense organ, not the soul; the Mīmāṃsākas, that the sacrifices produce a special kind of “power” (Śakti) that will ensure that the sacrificer (or the person who initiates it)16 attains the fruit of the sacrifice (Halbfass 1980, pp. 275–80).

Udayana briefly criticizes the Sāṃkhya view—a view anticipating behaviorism—that beliefs, desires, and actions are properties of the inner sense organ (something like the brain) rather than the soul (the soul being only a pure witness consciousness, rather than a doer or an enjoyer). And he criticizes as well the extreme dualism of the Sāṃkhya (Dravid 1996, pp. 69–75; cf. Sinha 1999, pp. 51–54; Joshi 2002, pp. 63–66).

His criticisms of the Mīmāṃsākas are far more elaborate. Those criticisms aim to show that the property produced by the sacrifices, a property that somehow links the performance to its fruit, is best understood to be a supersensible quality the soul acquires, indeed a supersensible merit, not the potency or power the Mīmāṃsākas posited. In this regard he first argues against the Mīmāṃsā causal doctrine in general, holding, for several reasons, that it would be better to posit universals to explain causal regularities rather than positing a separate category of causal power. He goes on to argue that the Mīmāṃsā doctrine posits more than is needed to explain the causal relation, and that reducing causal powers to certain essential and contingent qualities of substances encourages an empirical investigation of causal regularities in the way that positing a separate category of causal power doesn’t. And he responds to the main criticism the Mīmāṃsākas leveled against the Nyāya-Vaiśeṣika doctrine: he argues that relevant absences (e.g., an absence of wetness) should be taken to be necessary causal conditions of effects. Finally, he argues that (1) the scriptural passages the Mīmāṃsākas point to in support of their doctrine can be understood in a different way than they suppose, and (2) their independent arguments for the conclusion that certain inanimate substances are invested with a mysterious power rooted in the rituals give a supernatural account of certain causes when a natural account is available.17 In these ways he seeks to establish that it is more plausible to posit a supernatural quality, namely merit, as the link between the sacrifice and its fruit, than a supernatural power, preparing the way for the second argument of Stabaka V, an argument for the conclusion that only an omniscient, omnipotent being could ensure that finite souls are appropriately rewarded in accord with their merits (Dravid 1996, pp. 404–9; cf. Sinha 1999, pp. 265–74; Joshi 2002, pp. 234–35; Cowell 1864, p. 69).

4. The NK’s Second Stabaka

4.1. AGRV: The Argument for God as Revealer of the Vedas

In Stabaka II, Udayana builds on the conclusion of Stabaka I to argue that only an “extraordinary and omniscient being” (Dravid 1996, p. 106) could have revealed to human beings the means of obtaining merit by performing the sacrifices.

He tersely summarizes the dialectic of Stabaka II:

Since knowledge requires an external source, since creation and destruction take place, and since none other than He can be relied on, there is no other way open.18

The argument Udayana succinctly states in this verse can be detailed and formalized as follows:

• Given the efficacy of the Vedic sacrifices [their efficacy is established in Stabaka I, as explained in Section 3 of this paper], some means is available to human beings for knowing how to perform the sacrifices, and the possible means for our knowing how to perform the sacrifices are: (i) perception, (ii) inference, and (iii) word-based knowledge contained in the Vedas.
• Neither perception nor inference is the means by which human beings alive now know how to perform the sacrifices.
• If human beings alive now know how to perform the sacrifices through the word-based knowledge contained in the Vedas, then either (i) the Vedas themselves are the ultimate, eternal source of knowledge about how to perform the sacrifices, or (ii) not.
• The Vedas aren’t themselves the ultimate, eternal source of such knowledge.
• If the Vedas themselves are not the ultimate source of knowledge about how to perform the sacrifices, then either the Vedas were authored by an enlightened sage, or the Vedas were authored by an extraordinary, omniscient being.
• The Vedas were not authored by an enlightened sage.
• [So,] If the Vedic sacrifices are efficacious through adṛṣṭa, then an extraordinary, omniscient being authored the Vedas (Dravid 1996, pp. 105–207; Sinha 1999, pp. 71–116; Joshi 2002, pp. 82–130).

Though we will not here summarize all of the complex dialectic of Stabaka II, the parts we will summarize are sufficient to rule out the Mīmāṃsā view of the Vedas as authorless and eternal, as well as ruling out the Sāṃkhya view according to which they were authored by the sage Kapila, the semi- mythical founder of the Sāṃkhya school.

Udyana took it as obvious that people could not perform the sacrifices unless they know how to. Hence some source of knowledge must be available to them. Udyana further takes it as obvious that perception could not be such a means, especially since we cannot perceive adṛṣṭa. This leaves inference as a possible means of knowing how to perform the sacrifices. Udayana takes it to be so obvious that inference could not be the means of knowing how to perform sacrifices that he doesn’t even discuss that possibility. Nor did he need to, since all Hindus took it that the means of knowledge for performing the sacrifices can only be found, at least without a new revelation, in the Vedas. Compare this with the fact that Roman Catholics hold that the knowledge that priests have of how to perform the Mass must be traced, however indirectly, to the Bible.

Christians, however, hold that the Bible itself is an inspired text and that God himself is, in a sense, its author. The Nyāya held the same concerning the Vedas, but felt the need to establish that view, since the Mīmāṃsākas held the Vedas to be authorless and eternal texts, and the Sāṃkhyas held that the sage Kapila authored them. Udayana spends most of Stabaka II arguing against the Mīmāṃsākas.

Here we wish to detail Udayana’s argument for the view that the Vedas were authored by an apt person, i.e., one knowledgeable concerning that of which they speak, who has no desire to deceive. Establishing this is sufficient to show that the Vedas were not authorless or eternal, though Udayana gives a barrage of further arguments for the contention that there is no sense in which the Vedas could be held to be eternal.

Unlike western philosophy, classical Indian philosophy focused not on beliefs (understood dispositionally or non-concurrently) but on passing cognitive episodes. The general word for these was jñāna, a word which Matilal, whom we follow here, translates as “cognition”. There was a debate among Hindu philosophers about whether or not indeterminate perceptions (which they held to be a species of cognition) exist and, if they do, how they should be understood. An indeterminate perception is of a this here something now, but is not a perception of something as being an F. Such a “perception” always regards some substance, some of the universals it instantiates, and some of the qualities that inhere in it, but neither as related nor unrelated. One cannot be aware of such a perception because of its indeterminacy. Determinate perceptions are of things as characterized by universals and property tropes (e.g., of a blue pot as being a blue pot), and these are really distinct for the Nyāya from the things they characterize. Indeterminate perceptions, being sub-propositional, are neither true nor false, while determinate perceptions, i.e., perceptual judgments, are always either true or false. And we can have determinate cognitions of entities we cannot perceive, such as heaven, or Īśvara. The Mīmāṃsākas seem to have generally agreed with the Nyāya with respect to these points about perception and determinate cognitions, though they differed from them on certain details.

Some determinate cognitions can be called pramā. A cognition that is pramā is true, and further, so related to what caused it that it could not have been false or produced in the wrong way, though it might not have occurred, of course. Such cognitions are tied to reality in (nearly) the tightest possible way, even though those that characterize finite beings are caused. Matilal thinks the best translation of pramā in English is simply “knowledge”, with the qualification that the Indians were primarily concerned with occurrent knowledge, i.e., with knowledge episodes (Matilal 2002, pp. 149–53).

A key difference between the Nyāya and the Mīmāṃsākas, a difference that is important for understanding the dialectic of Stabaka II, is that the Mīmāṃsākas held that determinate cognitions are by nature pramā. It is like thinking that humans are by nature good—i.e., born with a full set of cognitive powers, a full set of legs and arms, etc. If a child is born deformed, that must be due to some positive defect in its causes. So too with cognitions. If they are not pramā, that must be due to some positive defect in their production. For the Mīmāṃsākas, all defects are positive in nature, and with respect to cognitions, include things such as diseases of the eyes or ears or, in the case of cognitions based on words, error or desire to deceive on the part of the person whose words ground the cognition.

The Mīmāṃsākas held that, since cognitions are by nature true, we have a right to believe them, nay should believe them, unless we have reason to doubt them—i.e., unless we are presented with a defeater for them. The Mīmāṃsākas applied this doctrine to word-based knowledge, what we would call knowledge based on testimony. Like the Nyāya (but not the Vaiśeṣikas), they held word-based knowledge to be a separate source of knowledge, one that can’t be reduced to inference.19 Any cognition we have based on words should be accepted as true unless we have a reason to think it is false. In the case of the Vedas, since they speak of transcendental or non-empirical matters the purpose of which is to be action-guiding, it is not possible for them to be falsified, and, since they are authorless, they cannot be held to have been written by someone in error, or with a desire to deceive.20

In contrast to the Mīmāṃsākas, the Nyāya-Vaiśeṣikas held that determinate cognitions are neither by nature true nor by nature false, just as human beings are neither by nature virtuous nor by nature vicious—which one they are depends on how they were raised, on the choices they make over time, etc. This means, for the Nyāya-Vaiśeṣikas, that a special excellence, a guṇa, is involved in the production of all cognitions that are pramā—lack of defects is not enough (Phillips 2012, pp. 23–32). But if such a special excellence is present, that will rule out both the contradictory and the contrary defects opposed to it. If a person has knowledge of some matter, she will neither be ignorant, nor in error with respect to that matter. This aspect of the Nyāya-Vaiśeṣika view is key to the part of Udayana’s dialectic we are focused on here.

4.2. Defense of Premise (4)

We turn now to that dialectic, a dialectic intended to support premise (4) of AGRV by showing the Vedas had to have been composed by an apt person. This dialectic begins with an argument against the Mīmāṃsā doctrine that cognitions are by nature pramā, i.e., instances of knowledge.

• If cognitions were by nature pramā then the general causes which are necessary and sufficient for the production of any cognition would be sufficient for any cognition to be pramā.
• If the general causes which are necessary and sufficient for the production of any cognition were sufficient for any cognition to be pramā, then all cognitions would be pramā.
• It is false that all cognitions are pramā.
• [So,] It is false that cognitions are by nature pramā (Dravid 1996, pp. 106–12; cf. Sinha 1999, pp. 71–77; Joshi 2002, pp. 83–89).

The Mīmāṃsākas denied premise (1) by insisting that cognitions that are not pramā are produced by certain positive defects, e.g., a diseased eye, in addition to the general causes productive of all cognitions. The Mīmāṃsākas wanted to say that the general causes in question must be “normal” and non-defective, and the generalization here should be held to be “for the most part true”, in something like Aristotle’s sense.

Udayana’s response to this objection is that defects need not be positive. For example, failure to observe differences between things, say two kinds of flowers, may lead to the cognition that they are the same in kind (certainly an instance of a cognition that is not pramā), and attending to such differences is something positive: it is a positive excellence that is opposed to a negative defect.

The Mīmāṃsākas admit this, but insist it holds only for perception. In the case of word-based knowledge, error and the desire to deceive, both positive defects, are the only things that could lead to a cognition that is not-pramā. Furthermore, since the Vedas are eternal and authorless, these positive defects cannot be thought to apply to them.

Udayana responds, first of all, by pointing out that cognitions based on inferences must not only be free of all fallacies (positive defects), but must also have certain positive excellences. He notes, for example, that though the absence of a deviant middle term (something analogous to an undistributed middle term) is necessary for an inference to be pramā-producing, more is required. It is necessary that the middle term be “in” the subject (or minor term) of the inference, that the middle term be invariably concomitant with the major term (and in the right way) and, finally, that the person performing the inference knows all this. And these are all positive excellences.

The Mīmāṃsākas, at this point, retreat to the ad hoc position that word-based knowledge is unique. In the case of word-based knowledge, absence of error and the desire to deceive are sufficient for a cognition’s being generated by words to be pramā: no positive excellence is needed.

Seizing on the ad hoc nature of this reply, Udayana argues that, if the Mīmāṃsākas insist on it, the Nyāya can, with equal right, insist that word-based cognitions that are not-pramā are not because of the absence of such excellences as knowledge or the desire to reveal what one knows to another who desires, or needs, such knowledge. The two views are counterbalanced. And, Udayana insists, it is not the case that the absence of positive defects is enough for word-based knowledge—the source of such knowledge must be in someone who himself has knowledge, knowledge that is not itself, ultimately, based on words or inferences. If you want medical knowledge you should go to a medical doctor, not a historian, even if the historian has no desire to deceive and is not in error about any of her historical beliefs.

The Mīmāṃsākas are now cornered, and forced to admit that word-based cognitions rooted in human testimony must be traced to a person who has: (1) knowledge (which rules out both ignorance and error), (2) the desire to tell the truth (which rules out the desire to deceive) and (3) the desire to impart his knowledge to someone who desires or needs it (which rules out the lack of such a desire). But they insist that none of these things holds of the Vedas, since the Vedas are authorless.

Udayana says that, even if the Vedas are authorless, that fact could not, by itself, give us any reason to think they are a source of knowledge. In other words, why should an eternal, authorless text (supposing such to be possible), simply in virtue of being eternal and authorless, be thought to correctly represent reality? And as noted, later in Stabaka II Udayana gives a barrage of arguments against the possibility of eternal texts.21

4.3. Defense of Premise (6)

We move now to premise (6), since premise (5) does not require proof—at least, Udayana believed it does not.

After his very lengthy refutation of the Mīmāṃsā doctrine of the nature of the Vedas, Udayana provides a refutation of the Sāṃkhya position that a human being, such as the sage Kapilla, the semi-mythical founder of the Sāṃkhya school, could have authored the Vedas by way of some sort of meditation or trance. Udayana’s refutation of the Sāṃkhya position is what provides support for premise (6) of AGRV (Dravid 1996, pp. 203–5; cf. Sinha 1999, pp. 113–15; Joshi 2002, pp. 128–30).

Udayana argues against the plausibility of this suggestion from the Sāṃkhya as follows: The Sāṃkhya hold that sages such as Kapila came to acquire the merit that allowed them to perceive adṛṣṭa by practicing dharma and engaging in the sacrifices; but this is circular—for how did even Kapila, absent divine revelation, know how to perform the sacrifices in the first place? Granted how complex they are, it is unlikely he came to perform them by accident, and even if he did, the mere performance of them would not have informed him of their power. And experience shows that no human being, simply by practicing yoga or engaging in meditation, comes to know of the existence of new kinds of realities utterly unknown to any human being before. Indeed, Udayana says that in this respect, trances induced by meditation are like dreams. No one who has never perceived or been told about a snake, for example, would first come to know what snakes are in a dream.

Nor could a person via meditation know that she will be rewarded in the next life by performing certain actions or living according to certain moral rules. All human perception, relying as it does on various sense organs, is of presently existing things. Of course, humans also have the power of memory, but memory is of things experienced in the past. And if the Sāṃkhya suggest that sages such as Kapila could know that performing certain sacrifices will produce good fruit in a future life because they remember performing them in a past life, Udayana would insist that we have no evidence that any human, unless aided by God, can remember specific things he did in a past life, much less that those things were the karmic causes of various pleasures and pains experienced in this life. Furthermore, this only pushes the question back. How did Kapila, in an earlier life, come to know of the very elaborate sacrifices that would lead to a better rebirth? If the Sāṃkhya respond that information from the Vedas could have helped him, such reasoning would be circular, since the Sāṃkhya are proposing that the Vedas themselves were authored by an enlightened sage.

5. Conclusion: The Broader Network of Argumentation in the NK

The argument that we’ve pulled from Stabakas I and II, which concludes that an extraordinary, omniscient being authored the Vedas, is not an argument for a big-G God. Furthermore, the argument in Stabakas I and II relies on accepting more than one claim that was controversial even in Udayana’s day, and in our own time there would be substantial and widespread disagreement with some of Udayana’s contentions. We ourselves take his key claim that the Vedic sacrifices are efficacious to be false, when understood in a standard sense (though a theist outside Udayana’s tradition might believe there is a non-standard sense in which it is true, if God hears all prayers of the pure of heart). But the reasoning in Stabakas I and II does not, of course, exist in isolation from the rest of the NK. The arguments Udayana develops in Stabakas I and II for the efficacy of the rituals, the existence of adṛṣṭa, and the divine authorship of the Vedas, are part of a broader network of philosophical arguments in the NK, a cumulative case for the existence of a big-G God who has revealed the Vedas, that reaches its apex in Stabaka V.

Given the focus of our paper, we limit our attention here to only a few of Udayana’s 18 arguments in Stabaka V (which he presents in two series, each with nine arguments), to call attention to aspects of this cumulative case particularly relevant to our focus.

The very first argument in Stabaka V, the argument “from products” or composite substances, may not seem obviously relevant to Stabakas I and II, but in fact, it nicely illustrates that just under the surface of Udayana’s argumentation throughout the NK, one finds interconnections that strengthen the weave of the cumulative case.22 Udayana’s celebrated defense of the argument “from products” is a sort of mirror image of his attack back in Stabaka I on Cārvāka naturalism. Now, in Stabaka V, Udayana argues that every product, i.e., composite substance, must have been made by an intelligent agent, and since some products could not have been made by an embodied intelligent agent, we must posit the existence of an unembodied intelligent agent as the maker of those products. In Stabaka I he had argued that the coming to be of the universe requires some cause(s) other than the inherence and co-inherence cause(s) of any effect, technically a nimitta (efficient/instrumental) cause, but in Stabaka I the nimitta cause he focused on was adṛṣṭa. Here the nimitta cause he focuses on is an intelligent agent who must be postulated as the maker of some composite substances, such as the Earth. The argument in Stabaka V builds on the causal theory Udayana defended at great length and subtlety in Stabaka I.

Jump now to the sixth, seventh, and eighth proofs in Stabaka V, which are centered on the Vedas, and so are more obviously relevant to our thesis. By this point in the NK, Udayana has considered God as creator, sustainer, and destroyer of all universes, and as exercising general providential care for living beings. He now turns to God’s special providential care, that is, his care for living beings with respect to spiritual and moral matters.

The sixth proof, “from authoritativeness”, argues from the Vedas having been unquestionably accepted by the large majority of people, generation after generation, to their having a divine author. This acceptance is a sign of their authoritativeness, but can’t be the cause of it, Udayana says. The cause of such certainty, as we have seen in discussing the dialectic of Stabaka II, can only be found in qualities of its author—above all, in its author’s omniscience.

That the content of the Vedas bears witness to their having a divine origin is brought out specifically in the seventh proof, in which Udayana says that the Vedas are composed by an omniscient person because they have “the nature of statements that have been accepted by a large mass of people while no other basis (besides themselves for their acceptance) is perceivable.” (Chemparathy 2018, p. 98). The second part of this conjunction rules out that our acceptance of the Vedas is what makes them Vedas. We don’t know of the nature of adṛṣṭa, or actions productive of it, except through the Vedas, and so, human knowledge of the world that is independent of the Vedas cannot provide any fundamental basis for accepting the Vedas as revealed. Furthermore, the authority of the Vedas cannot be traced to texts of the tradition based on them (e.g., the law code of Manu, or the Bhagavad-gītā) for these get their authority from the Vedas. Finally, though the scriptures of the Buddhists and the Jains speak of supersensible matters, they cannot be authoritative since they are not accepted by “a large mass of people”.

The eighth proof in Stabaka V offers another reason for thinking the Vedas were composed by an intelligent being: they consist of sentences, and all sentences are composed by an intelligent being. This is Udayana’s most elegant refutation of the Mīmāṃsā notion that the Vedas are authorless; it is rooted in the Nyāya-Vaiśeṣika doctrine that causality (understood to include agent causality) is regulated by universals, a doctrine Udayana defended at length in Stabaka I, and elsewhere in his works. Udayana goes on to say that the content of the Vedas shows their author must be omniscient. The proof mirrors the first proof of Stabaka V, the proof “from products” or composite substances. Here Udayana argues that every text, every collection (composite) of sentences, has an author, but that the Vedas, in virtue of their content, have a superhuman, omniscient author.

We have argued that Udayana reasons from the very content of the Vedas to their being revealed texts, and thence to the texts having been authored by a God-like being. Even if one accepts this interpretation of the overall line of reasoning in the NK, one might be disappointed by aspects of Udayana’s argument. Both in Stabaka I, and in the three arguments focused on the Vedas in the first series of arguments in Stabaka V, he emphasizes the long-held belief of the great number of people in the efficacy of the Vedic sacrifices and the authority of the Vedas. But these seem to be external matters, and though, as we have seen, Udayana also argues that the sacrifices are in fact efficacious and appeals to Vedic instructions (internal content) re. performing sacrifices, one might wish that he had said more about other features of their content. He might, for instance, have spoken of how the Vedas have been resilient through the centuries, or of how they meet spiritual or moral needs, or of their sublimity, etc. We understand this kind of complaint. But we think that Udayana indirectly addresses it, especially if one considers the second series of proofs in Stabaka V, which point to Vedic content evincing the truth of the Vedas.

In the second-series proofs Udayana gives an argument “from explanation” according to which only one who understands all of the Vedas both in their relation to each other and to what can be known by other means of knowledge of the world could possibly have explained the Vedas so as to make them resilient through the centuries. Udayana thinks the essential doctrines of the Vedas have been marvelously stable, including with respect to their elaborate rituals (despite disagreements regarding minor details). He argues that such a resilient tradition must have been grounded in the revelation of an omniscient being who could explain the Vedas in all their internal and external complexity.

Further, in the second-series proofs Udayana suggests that the Vedas fit our spiritual and moral needs. The ancient sacrifices help wash away or atone for sin, and the greatest of the sacrifices, one that called forth extraordinary effort, was aimed at attaining to heaven in an afterlife. In the second series of proofs, Udayana notes that the Vedas everywhere say God should be worshiped for his own sake, out of gratitude and love (Chemparathy 2018, p. 126), and in the final prayer he longs to see the Lord and be conformed to his goodness and perfection (Dravid 1996, p. 495; cf. Sinha 1999, pp. 342–43; Cowell 1864, p. 85). And in the second series-proofs we find a long argument (the longest of Stabaka V’s 18 arguments for God) that the moral injunctions of the Vedas, particularly their absolute moral prohibitions, can only be understood as coming from a flawless being, indeed a big-G God, who, like our fathers, always wishes us good (Dravid 1996, pp. 433–85; cf. Sinha 1999, pp. 297–334; Joshi 2002, pp. 248–56; Chemparathy 2018, pp. 122–25).

Finally, with respect to sublimity of teaching, Udayana chose to show rather than tell. He quotes passages from the Vedas and the tradition (especially the Bhagavad-gītā) throughout the NK, and abundantly in Stabaka V, including the second-series proofs. And these passages are sublime, which explains why the Vedas, and other works of the tradition, especially the Gītā, have commanded the respect and, indeed the love, of many who are not Hindu. There is something in their content that is transcendent, and uplifting, and seems to point to a divine source. Udayana also shows the sublimity of the Vedas in the prayers that conclude each chapter. They are profoundly moving and deeply informed by the Vedas and the Hindu tradition. We think they are part of Udayana’s case: he had said that his treatise concerns the supreme being “whose worship the wise consider the means to heaven and liberation” and that his entire treatise is meant to be an act of worship, the sort of worship only a “logician” is able to give. Udayana moves from argument about God and his existence, to worshiping God, and invites us into that worship. Given the title of the work, we think the prayers that conclude each chapter, especially the prayer that concludes the final chapter, should not be taken as mere ornaments, mere rhetorical flourishes, non-integral to the work—indeed, we think one cannot understand the work without looking to the prayers as keys to understanding each individual chapter, and that the final prayer must be taken into account in understanding all of it. The case for this comes from reading the prayers that conclude the chapters. Even in translation they are beautiful; they also express, concisely and with great poetry, Nyāya doctrines concerning God.

Udayana crowns the second series of proofs by pointing to passages in the Vedas that teach that Īśvara has the properties of “omniscience, contentment [goodness], eternal knowledge, independence, a power that remains undiminished at all times, a power that is endless”; in this way he argues that the Vedas themselves teach that a big-G God exists (Chemparathy 2018, p. 119; cf. Dravid 1996, p. 429).

We have now made the case that the NK should be understood as a classic instantiation of what we might call “philosophical revelationism”.23 Philosophical revelationism is not any species of fideism, but neither is it a species of standard natural theology. It is an approach to natural theology that relies on the content of a putative revelation early in the process of arguing for the existence of a divine revealer: at the same time that the plausibility of a putative revelation is established, in part through focus on its content, the existence of a divine revealer is established. Udayana constructs an inference to the best explanation24 for the conclusion that the efficacy of the sacrifices points to the reliability of the Vedas and the existence of a divine revealer, and then, building on this, develops multifaceted argumentation that the revealer of the Vedas is a big-G God.

Worry may remain about circularity. Even Chemparathy, who among all the English-language commentators on the NK has the interpretation closest to ours, seemed to strictly separate philosophical arguments for God’s existence from arguments that bring the content of putative revelation into the database. He says that the non-Vedically based proofs in the NK seem to assume certain doctrines based on Hindu scriptures and thus to be theological rather than philosophical in nature (Chemparathy 2018, p. 154). And he thought the second-series proofs were simply meant to refute the Mīmāṃsākas, and to have no independent worth in arguing for a big-G God.

But a charge of vicious circularity will not hold against Udayana’s strategy. It is true that the claim “There’s a divinity who has revealed” embeds the claim that “There is a divinity.” But the embedded claim does not need to be established before the more complex claim. If the SETI (Search for Extraterrestrial Intelligence) project finds intelligent life outside our planet, researchers and analysts will be saying something like: “The radio signals SETI detected could only come from intelligent beings outside our planet.” That claim embeds the sub-claim “There are intelligent beings outside our planet”, but we might come to know that simultaneously with knowing the embedding claim.25 In the NK we have an inference to the best explanation for a divine revealer that depends on the content of the Hindu revelatory claim, and at the same time that the (likely) existence of a divine revealer is established through this IBE, the plausibility of putative revelation is established. Because this enterprise does not assume content of putative revelation to be true, it is natural theology. But it is not SNT, not standard natural theology, as we have defined it, because the content of putative revelation is part of the database under consideration.

Udayana’s approach can be generalized: persons of any faith who appreciate the centrality of their holy scriptures can bring the content of those scriptures into the database that grounds explanation—the content can be brought in as putative revelation, though not, in a philosophical argument for God, as actual revelation.

The possibility of generalizing Udayana’s approach points, of course, to the existence of different religions, with different ideas about what has been divinely revealed. This raises an important question, though not one we will pursue here (at any length): Is the diversity of religious traditions problematic for the approach we’ve ascribed to Udayana and are championing? From the standpoint of contemporary philosophers drawn to the Judaeo-Christian tradition, whether one takes it to be problematic may depend partly on how one understands the contention in Hebrews 1:1 that God spoke “to our ancestors in many and various ways”.26 And while we cannot know exactly how Udayana would address what is now called “the problem of religious diversity”, he provides hints in two passages of the NK. The first raises the objection that there is no need to enter into logical investigation of the Lord’s existence since all sects and persons worship the Supreme Being under some name or description (Dravid 1996, pp. 4–5; Sinha 1999, pp. 4–6; Cowell 1864, pp. 2–3; Chemparathy 2018, pp. 81–82). Udayana here mentions not only different Hindu schools of thought, but also, for example, the Buddhists, who he says worship God as “the Omniscient” (Cowell 1864, p. 2). In responding to this objection Udayana does not deny that all sects in fact worship God under some name or description, but instead insists that logical investigation of God’s existence can be considered a form of worship that follows on hearing the scriptures. Later commentators illuminated this passage (Sinha 1999, pp. 7–9) by pointing out that some who worship God may not realize they do, and that others may have false conceptions of God. Udayana concludes his list of those who worship God in some way by saying: “the followers of the Nyāya [worship God] as Him who is all that is said worthy of Him,” and goes on to identify God as Śiva (Cowell 1864, p. 3). The second passage in the NK that hints at an approach to religious diversity, which occurs right before the final prayer, quotes chapter 9 of the Bhagavad-gītā: Kṛṣṇa, an incarnation of the Lord, says he will receive the prayers and worship of all those whose faith and love for him is sincere, even if they worship him under the name of other gods, or are in some ways mistaken about his nature and attributes (Dravid 1995, pp. 494–95; Sinha 1999, pp. 341–42).

Udayana reminds us not to restrict the database in arguing for the existence of God: we should not ignore data from the content of putative revelation. That lesson itself may be writ large: our explanations should be part of the most comprehensive explanation possible, one that subsumes every branch of philosophy, and every aspect of reality. All explanations or systems have parts that look problematic; in the end, one needs to undertake comparative analysis among the available options and pick the overall best. Udayana, like all great philosophers, was a consummate systems-builder, though one who, as he would admit, relied greatly on a long tradition. It is a testament to the tradition that it has endured many centuries and has withstood repeated assault. It has modified its teaching, but in ways that keep its central doctrines intact.

The Sāṃkhya school of thought is now basically dead, at least as an independent school, its most enduring insights having been absorbed long ago into the Advaita Vedānta. The Mīmāṃsā school, in virtue of its rules for interpreting texts, rules that are neutral with respect to the content of the texts interpreted, remains alive in the thought of traditionally trained scholars of the Vedas and Hindu law, and its subtle theory of language still commands respect from those who have studied it and who are concerned with some central problems of philosophy of language that its greatest philosophers addressed with skill and insight. But when it comes to what the Mīmāṃsā school cared about most—the reduction of the Hindu religion to performing moral and ceremonial duties—it lost, long ago, the war it bravely waged for centuries. The writings of Vedāntins, such as the great Śaṅkara himself, played a role in this; but so, too, did Udayana’s NK.

The Nyāya-Vaiśeṣika school, by way of contrast, remains alive, and philosophers trained in it have written works that command the attention of 20th and 21st century philosophers, all of whom would acknowledge Udayana’s power in keeping the school alive through a time of great controversy, and the continuing relevance of his synthesis of the Nyāya and Vaiśeṣika schools, a synthesis accomplished during a time when all Hindu schools were attacked by Buddhist logicians wielding incredibly subtle arguments. Udayana repelled this subtle philosophical onslaught by fashioning new intellectual tools and engaging in nuanced and remorseless, but fair and honest, attacks on Buddhist doctrines, all the while building up a magnificent intellectual edifice in defense of Nyāya-Vaiśeṣika doctrines, an edifice he crowned with his Nyāyakusumāñjali, his Flower Offering of Logic, ending it with a profound prayer of praise to the Lord. Metaphysics, epistemology, ethics, God: they lock together into a lovely whole, greater than the sum of its parts.