Universal Properties in Indian Philosophical Traditions
UNIVERSAL PROPERTIES IN INDIAN PHILOSOPHICAL TRADITIONS
Early Grammarians on Universals of Words and Meanings
In ancient India systematic metaphysics started with a linguistic turn. Ontological concepts and controversies arose in the context of musings on meanings of words and debates on declensions, unlike in ancient Greece, where metaphysics arose out of wondering about numbers, figures, and nature. In Pāṇini's grammar and his early commentaries (between the fourth and second centuries BCE) the three crucial technical terms for a universal—sāmānya, jāti, and ākṛti —were already explicitly in use. Philosophers of language dabbled in metaphysics since Patañjali's "Great Commentary" to Pāṇini's grammar. The device of adding a tva or tā (roughly equivalent to the English "ness") to any nominal root x, yields, as meaning, the property of being x. From substance (dravya ) one can thus mechanically abstract substance-ness (dravya-tva ), from real (sat ) and reality (sattā ). With this device in place it was natural to make the distinction between an individual substance and the property that makes it what it is, its abstract essence. But even to parse this talk of concrete cows rather than of the bovine essence, the grammarians drew the distinction between talking about one particular cow and talking about any cow or a cow in general (VMB on Pāṇini sutra 1.2.58 and 1.2.64). The distinction between the general and the particular also came up for discussion in the context of the logic of pluralization. What allowed one to say "trees" or "men" instead of using the word for a tree or man as many times as the number of trees one referred to? It must be because the direct meaning of a common noun is the shared universal property of the referents that one could eliminate all but one remaining occurrence of that word, when speaking generally. One could also issue universalizable moral imperatives such as, "A cow ought not to be killed," which, Patañjali jokes, is not obeyed by simply sparing the life of one single cow.
Jāti (a word that, in modern Indian vernaculars, has come to mean "a class," "a caste," or even "a nation")—the Sanskrit counterpart of the Latin "genera"—is used by Pāṇini for a shared property of all the particulars of one natural kind, which serves also to distinguish any one of them from things of other kinds. The particulars are called vyakti —a word that etymologically suggests a distinct concrete manifestation of common and uncommon properties. The problem with this universalist theory of meaning—defended by Vajapyāyana—was that when, in a descriptive or prescriptive sentence, the action denoted by the verb has to hook up with what the noun means, it has to be a particular. For, after all, no one can bring cow-ness, cut the tree-essence, or meet humanity on the street.
Thus, in Indian philosophical semantics the dispute between those who insisted that a word primarily means a universal and their rivals who held that it must be particular substances that are the first meanings of words is at least twenty-two centuries old. The word often used for universal by Patañjali was ākṛti (literally "shape"), which is more reminiscent of form than a property. In answer to the basic question "What is a word?" Patañjali considers the option, "Is it that which remains non-distinct among distinct individuals, un-torn when individuals are torn down?" and answers, "No, that is not the word, that is only the universal (ākṛti ).
The need to switch to imperishable universals as meanings was felt both by the grammarians and the Mīmāṃsā school of Vedic hermeneutics for whom the authority of authorless sentences of the Vedas rested on their eternity. The relation between words and objects was said to be entrenched and eternal. If perishable particular horses, cows, humans, and plants were the meanings of words, how could they be the eternally connected meanings of these beginningless Vedic words? The word gauḥ (cow) is therefore best taken to be eternally connected to the timeless bovine essence.
The first clear recognition of the need to postulate universals might have come, not so much from the theory of meaning but from reflecting on the generality or repeatability of the audible words themselves. That there could be many pronunciations or distinguishable phonations of the same word was seen to be an unquestionable example of the one-in-many. That naturally went hand in hand with the idea of the real word-type existing timelessly there independently of its temporal perishable token-utterances. Later, in the philosophy of Bhartrhari, sometimes called a linguistic nondualist, word-universals and meaning-universals and one's natural tendency to superimpose the former on the latter were elaborately discussed, because it was easy to confuse them with Bhartrhari's single most important metaphysical concept of a speech-bud or linguistic-potentiality (sphoṭa ) in all consciousness, where signifier and signified exist undivided, waiting to blossom into articulated structures of sentences.
In the context of interpreting Kātyāyana's aphorism, "the word-meaning-relation being fixed," Patañjali mentions two alternative ways of taking the concepts of form (ākṛti ) and content or substance (dravyam ). In the first sense forms are universal properties that remain unchanged while individual material substances come and go, hence the forms must be those fixed meanings. In the second sense, somewhat like René Descartes's lump of wax, the substance continues to exist, retaining its sameness while the structures vary or perish, hence the substance or content must be that fixed meaning. If one defines the universal as the invariant across variations of individuals, then that definition fits both the form (under the first interpretation) as well as the indestructible content (under the second interpretation). One encounters a similar clash of intuitions in Ludwig Josef Johann Wittgenstein's Tractatus Logico-Philosophicus, where, about the ultimate constituents of all atomic facts, one finds the remark, "Objects are form and content." This idea of the enduring stuff of changing entities as a ground of sameness, found in early grammarians' and Advaita Vedāta thought, was later on picked up by the Jaina notion of a vertical universal (ūrdhvata-sāmānya), as against the more common property-universal that was termed horizontal universal (tiryak-sāmānya).
The Hot Topics for Debate
Between fifth and fifteenth centuries the debate between mainstream Nyāya-Vaiśeika and Mīmāṃsā realists and Buddhist nominalists raged around the existence of eternal essences. The major points of disputation were:
(1) Must one explain the use of a common noun or the experience of community across a plurality of particulars by postulating a single real property inherent in each of those particulars? (Vaiśeika and Mīmāṃsā said yes with some caveats, and Buddhists said no.)
(2) Is this property totally distinct from the individuals that exemplify it? (Vaiśeika said yes, and Bhāṭṭa Mīmāṃsā said yes and no.)
(3) Does a universal exist only in all its own instances or are universals omnipresent? (This is a trick question set up by the Buddhist nominalist, answered cautiously by Vaiśeika.)
(4) Do universals have any role in causation? (Vaiśeika said that they can cause one's awareness of them. For Buddhists anything that is eternal must be causally barren, hence nonexistent. For Udayanācārya [tenth-century Nyāya-Vaiśeika] nomic relations of necessary concomitance are ontologically founded on the universals inherent in the causes and effects.)
(5) Can the work that is done by universals be done by relations of resemblance between particulars? (Vaiśeika said no, Jainism and Madhva Vedāta said yes).
The Classical NyĀya-VaiŚeika Realism about Universals
Universals come to occupy a crucial role as the fourth type of real, in the scheme of six basic categories of reals or "things-meant-by-words" (padārthas )—notice again the semantic orientation—listed in the Vaiśeika sutras of Kaṇāda. In that canonical scheme, after the three types of unrepeatables—substances, particular qualities, and motions—come common properties. Although substances, qualities, and motions are entities of different types, they share one common property: They are all real. What is this realness that is common to all substances, qualities, and motions? Realness is a generic essence present in many substances, qualities, and motions. It is a universal, the highest one. Then there are less general features as well, the substance-hood shared by all substances, the quality-hood common to all qualities, and the motion-hood inherent in all motions. These second-tier universals are called common-uncommon since they function as defining properties belonging to all the members of the class to be defined, and lacked by all else.
The Vaiśeika sutra's word for universal is sāmānya " (the phonetic resemblance with "sameness" may not be entirely accidental), meaning "what is common." The word for an individuator or particularity is víseṣa, which means "uncommon feature" or "specialty," the difference-maker. Flower-ness could be a common property, shared by roses, jasmines, and sunflowers. But the same property would be a difference-maker when you compare a rose with fruits, seeds, stones, and animals, since none of these except the rose has flower-ness. Hence, Kaṇāda's aphorism, "Universal and particularity depend upon understanding" (VS:1/2/3).
Commentators hasten to point out that this formulation does not mean that universals are subjective or invented by one's ways of understanding the world. All it means is that one finds out by the verdict of one's understanding whether some property is a pure universal or also a demarcator, as shown earlier.
Four broad arguments are generally proposed by these staunch realists for proving the existence of universal properties:
(1) The evidence of sense-perception is the strongest of all. Unless one is threatened by a logical inconsistency, one must admit some common recurrent entity in each of those many things that sense-perception shows one to be of the same kind. This class-character, the basis for one's sense of sameness (anugata-pratīti ), is a universal.
(2) The argument from the meaning of general words runs as follows. A learnable common noun such as bird can denote an unlimited number of particulars of enormous variety. How the same word with the same meaning can correctly apply to so many diverse particulars calls for an explanation. The explanation must lie in a distinction between reference (śakya ) and sense (śakyatāvacchedaka ). Thanks to the existence of an objective universal, for example, bird-ness, which serves as the same sense, the same word can distributively refer to all birds or any bird. This does not boil down to one of the early extreme views that the bare particular or the pure universal is the primary meaning of a word. It is the balanced view that the meaning of a word is a particular possessing a general property that serves as the common mode of presentation of its unlimited number of referents.
(3) Then one has the argument from lawlike causal connections. Fire is a substance, but when it causes burning, its causal efficacy is not determined by its simply being a substance, for, then any substance would burn. To explain what makes fire—and not any other substance—the cause of burning, one needs to postulate fire-ness as the property that limits the causality of fire toward this effect. With the advent of extremely technical New Nyāya (around the thirteenth century) the need to have limiters (avacchedaka ) of cause-hood and effect-hood became the standard ground for ontological commitment to universals.
(4) Admission of universals also helped Nyāya solve the problem of justifying the inductive leap from observation of a few cases to a universal generalization covering all cases of a concomitance. The common property observed in a few instances can, as it were, put one in direct perceptual touch with all the other instances where also it inheres, not in their individual details but in a generic way. Here, the universal itself is supposed to play the role of the operative connection between the sense-organ and the apparently unobserved instances of that universal.
With all these supporting arguments for its existence, the precise definition offered by Nyāya-Vaiśeika settled down to this, "A universal is that which, being eternal, is inherent in many." Not any quality inhering in a substance is a universal. A wish inheres in a soul, but it is a short-lived episode, not a universal. Colors are not universals in this system because they are unrepeatable qualities clinging to the particular surfaces. All colors share the universal color-hood. But two red apples have two distinct red colors in them, just as each of them would have a distinct falling-motion when they both fall. A universal must subsist wholly in each of its instances by the special relation of inherence. A universal must be wholly inherent in each of its instances. The word inherent must be taken seriously. A single string may be running through many flowers, but it is only in contact with them, the whole string is not inherent in any one of them.
What is inherence? It is a kind of being-in, the converse of which is an intimate "having." Humanity inheres in me, just in case I have humanity. Now, having can be of many kinds. Things have qualities and motions. Wholes have parts. I have a pen in my hand. A rich man has a big house. The logical structure of each of these relations of characterization, constitution, contact, and ownership, however, is utterly different. All four are more or less aptly reportable by the use of the preposition in or of : the taste is in the apple, the room is or consists in the walls, roof, and floor, the pen is in between the fingers, and the house is of the rich merchant. Still, one initial grouping could be made to clarify their distinct structures. The taste and the room cannot exist without the apple or the room-parts. The taste cannot float about on its own, minus the apple. The room cannot stand independently of the walls. But that pen can easily exist untouched by the hand, and that house can change hands.
So, the first two relations hold between pairs that are "incapable of standing apart from one another" (ayutasiddha), whereas the other two relations hold between pairs that are "capable of standing apart from one another" (yutasiddha). However tightly my ring is stuck to my finger, it is not inherent in it as inseparably as finger-ness is inherent in my fingers. It is no physical glue but a metaphysical inseparability that joins the goat-ness to the goat, ties up the running and black color of the goat to the goat, as well as binds the goat to its body-parts. The kind of being-inseparably-in that connects the universal to its instances has to be distinguished from the way a berry lies in a bowl. For the sake of economy—the principle of not multiplying entities beyond necessity—the mainstream Nyāya-Vaiśeika metaphysicians posit only one single such relation as enough to link innumerable pairs of universals and particulars, qualities and substances, and wholes and parts. For systemic reasons, this relation is supposed to be eternal as well. And this is inherence (samavāya ). Even other universal-friendly realists, such as the Bhāṭṭa Mīmāṃsaka, give Vaiśeika a lot of grief over this peculiar theory of the exemplification. The Bhāṭṭas themselves take the relation between a universal and its own exemplifier to be identity-in-difference. The Buddhist logician finds both inherence and identity-in-difference equally unpalatable.
Though one cannot experience Vaiśeika universals by themselves, they are ontologically independent of the particular instances. Even when all cows are destroyed in the world, cow-ness will still be around, for otherwise the possibility of a fresh cow coming to be remains inexplicable.
Real Universals and Titular Properties: On Being a Cook
Though all universals are common features, not all common features corresponding to multiply applicable descriptions are, strictly speaking, universals. Being a Brahman (a member of the highest priestly intellectual class) is taken to be a natural kind by Nyāya-Vaiśeika in the face of vehement opposition by anticaste Buddhists and Jainas. But being a cook is the standard example of a common feature that is not a real universal. The Nyāya-Vaiśeika philosophers suggest six tests that an alleged (semantically suggested) property must pass to count as a genuine universal. These tests or hurdles are called universal-blockers:
(1) If a property has only a single exemplifier, then it is not a universal. "Being the Statue of Liberty" is not a universal, neither is time-hood, because there is no more than one Statue of Liberty, one time.
(2) If two properties have exactly the same extension, for example, the property of being a Homo sapiens and the property humanity, they cannot be two distinct universals.
(3) The domains of two universals can be either completely disjoint or one of them completely included in the other. They cannot be partially intersecting and partially excluding each other. Thus, being material and having a limited size cannot both be universals in Vaiśeika ontology, because while lots of things have both the properties, open space is supposed to be material yet not limited in size, while the internal sense-organ is supposed to be limited in size but immaterial. Whether crosscutting disqualifies both the properties or only one of them, and whether the neat ontological hierarchy that is presupposed by this universal-blocker is integral to a realist metaphysics have been the subject of much contemporary debate (see Shastri 1964, Mukhopadhyaya 1984).
(4) A regress-generating property is not a universal. Universal-hood is not a universal, although all universals seem to have that property in common. Because then one could multiply levels of universals endlessly. Universals do not have further universals in them.
(5) When the nature of a characteristic is to merely distinguish its bearer, for example, one earth-atom, from another particular of that kind, such ultimate individuators should not be brought under a general category of individuator-hood, for that militates against their necessarily unique nature. Failing this test, the alleged generality individuator-ness (visesatva ) fails to qualify as a universal within Vaiśeika atomism.
(6) The feature must bear inherence and no other relation to its bearer. Inherence-hood is not a universal because, had it been one, it would have to be related by inherence to inherence, which would be absurd. An absence cannot be a universal. Nor could the negativity common to all absences be a universal. Even though every rabbit is hornless, neither the absence of horn itself nor the absence-ness of the absence resides in rabbits or absences by inherence. Besides these, compound properties such as being a sturdy black cow or being either a cow or a buffalo are ruled out because universals are supposed to be simple.
What happens to the properties that, thus, get disqualified by a universal-blocker? They are thrown into the mixed pile of titular, surplus, or imposed properties (upādhi ). They could still be of much theoretical and practical use. Not only nonnatural generalities like being a New Yorker, but even is-ness, knowability, and positive presence (shared by items of all the six categories—substance, quality, motion, universal, inherence, and final individuator—but not found in absences) are merely titular properties. Knowability and existence (is-ness) are (intensionally) distinct properties, in spite of being equi-extensive, because they are not universals.
How are Universals Known?
One needs philosophical reasoning to grasp such deep universals as substance-hood, because many instances of substance-hood, such as time, atoms, other people's souls, are not objects of perception. If the instances are perceptible, the universals must be directly perceptible as well. One sees flower-ness in a flower, just as one sees its hue and smells its fragrance. According to Nyāya epistemology, to see Black Beauty as a horse one must first see its horse-ness (which is a perceived universal, though it is not perceived to be a universal).
But many strong arguments could be given against the perceptibility of universals (NM, ch VII). The following are a couple of examples:
If properties were perceived, one would perceive them even at the time of encountering the first exemplifier, but one does not. Hence properties are abstracted, not seen. Both the premises of this argument, of course, could be questioned. For the empirical knowledge of a common property to dawn gradually, a recognition must take place in the second, third, and subsequent sightings of the instances. To be faithful to the form of that recognition, "I have seen this sort of animal before," is to admit that even in the first instance that sortal property was seen.
Here is another antiperception argument. If properties were objects of perception, they would be causes of perception, but they are not. Therefore, they are not perceived. Again, both the premises are rejected by the Nyāya realists. Pot-ness need not itself reflect light back into the retina for it to be causally relevant to the visual perception of pot-ness. As long as the pot in which it inheres is in contact with the seeing eyes, it has a causally operative connection with the appropriate sense-organ. If, of course, perception is defined as prelinguistic and nonconceptual (as some Buddhists have done) and universals are taken to be word-generated concepts, then to use that definition as an argument for imperceptibility of universals would be crudely question-begging.
With Fregean sensibilities one could propose another quick argument against the perceptibility of universals. Universals are not objects but functions. Therefore, they are not objects of perception. Still, there is a clear shift in the meaning of "object" between the premise and conclusion of this argument. There is a basic (rationalist?) resistance even among realists in the West to admit sense-perception of universals, because universals are supposed to belong to the intelligible realm. In The Problems of Philosophy Bertrand Russell claims that one has direct acquaintance with universals, but that acquaintance is not meant to be sensory. It is only David M. Armstrong, whose view about universals comes close to Nyāya-Vaiśeika realism, who seems to have warmed up to the idea of perceiving universals.
Attacks from the Buddhist Nominalist
Vaiśeika's first argument for the existence of universals depends on the generalization, "In every case, the sense of commonness or similarity felt by word-users must be spawned by an objective universal." Surely, this generalization is riddled with counterexamples. One has just seen earlier how people feel a sense of similarity across many cooks, yet the Nyāya-Vaiśeika realists refuse to admit cook-ness as a universal. There is no good reason to posit these weird entities, and every reason to eliminate them. So claimed the Sautrāntika-Yogacara Buddhists, "It does not come there (from another place), it was not there already, nor is it produced afresh, and it has no parts, and even when it is elsewhere it does not leave the previous locus. Amazing indeed is this volley of follies!" (PV 1.152–153).
With this oft-quoted remark Dharmakirti (1994) summarizes his battery of objections against the Nyāya-Vaiśeika theory of universals. How can a universal remain the same while existing in distinct things and places? Does it scatter itself into parts or does it live in its entirety in each instance? When the locus moves, does it move? If cow-ness is everywhere, why is it absent in a horse? If it is only where its instances are now, then how does it travel to a new place when a cow is born there? It does not pervade the place where an individual is located, for then the place itself would be its instance, yet how can it manage to inhere in the individual that occupies that place? If the particular instance is needed as a revealer of the ubiquitous universal, how come one cannot perceive the cow—its revealer—independently of noticing the universal cow-ness? A lamp reveals the preexistent pot in a room, but one does not need to see the pot first before one notices the lamp (PV 156).
Most of these difficulties, the realists retorted, suffer from a category-mistake. They assume that a universal is just another kind of super-particular. But a universal is not a spatiotemporal thing, and that is why multiple-location without divisibility is not a problem for it. In spite of such robust responses Buddhist antirealism about universals became more trenchant in the second millennium until such caustic were remarks directed at the Vaiśeika realists, "One can clearly see five fingers in one's own hand. One who commits oneself to a sixth general entity finger-hood, side by side with the five fingers, might as well postulate a horn on top of one's head."
Apoha Semantics: The Buddhist Exclusionist Account of Concept-Formation
Buddhist logicians have an error-theory about universals and permanent substances. There are nothing but momentary quality-particulars in the world. But the human mind, afflicted by recurrence-wishes and language-generated conventional myths, has a tendency to cluster some of them together first in the fictional form of enduring substantial things and then further classify these "things" into types. This illusion of generality, of course, has some pragmatic value, because, except in contemplative experience, most of one's working cognitions of the world take the form of predictive or explanatory inferences on the basis of these apparently general features and their mutual connections.
When a particular cow (which is a fictional cow-shape superimposed on certain packets of quality-tokens) is seen to be other than all other animals, the original indeterminate (concept-free) perceptual content somehow causally triggers off this difference-obliterating tendency. The particular cow-image is made to "fit" this linguistic and imaginative exclusion from the complementary class of horses, rabbits, pillars, and such things. The specificity of the particular cow—its numerical detailed differences from other cows—is ignored; instead, this mere exclusion from noncows is foisted on to the perceptual content as a predicate. This exclusion masquerades as the universal cow-ness. To take Dharmakirti's (1994) example, the universal antipyretic-ness is a useful figment of imagination. In the external world there is no single shared intrinsic property of different medicinal plants all of which work as fever-reducers, except that they are other than those things that fail to relieve fever. Antipyretic-ness is an erroneous reification of this mere exclusion (apoha ). This, in a nutshell, is the apoha nominalism of the Yogacara Buddhist logicians.
Milder Nominalisms: Resemblance Theories
In the middle of this great battle between the realists and nominalists, the Jaina syncretists step in with the reconciliatory message that every object of knowledge has an alternatively more-than-one (anekānta ) nature—particularity and generality are just two of them. One cannot doubt that things do objectively resemble each other. These resemblances are real relations. But both the things and their mutual resemblances are particulars. Nothing has the burden of being repeatable.
The Jainas reject the Buddhist version of nominalism, more or less on the same grounds as Kumārila Bhāṭṭa, the great Mīmāṃsaka, rejected it. Positive predicates, Kumārila had objected, cannot all be given a negative meaning. Since these exclusions are nonentities invented by erroneous imagination, to say that all one's words mean them is to turn all words into empty terms. Indeed, since all exclusions are equally hollow in content, distinguishing one from another would be like trying to distinguish one imaginary nonexistent from another. Only those denials make sense that have something positive to deny. Since all descriptions capture only negations, this theory, ironically, strips one's negations of all meaning, since there is nothing left to deny.
Jaina thinkers reject the exclusionism of the Buddhist but use the Buddhist criticisms to reject the Vaiśeika realism. In its place they propose this resemblance theory. Prabhācandra anticipates the Russellian objection that at least all these resemblance-relations would ultimately need a shared resemblance-universal. His answer to it is that, just as a Vaiśeika final individuator (víseṣa ) does not need another distinguisher, one resemblance does not need a higher level resemblance or universal to explain why all those resemblances are similar. While accounting for the similarity between ground-level particulars, they also account for their own similarity to each other. Versions of this theory were adopted by followers of Rāmānuja (qualified monist Vedāta) as well as by Madhva (dualist Vedātin) logicians. Vyāsatīrtha of the latter school clarified how a single resemblance can reside, as it were, with one leg in the resembler and with another leg simultaneously in many other similar particulars.
The category of resemblance admitted by these philosophers is different from the resemblance admitted by Prābhākara Mīmāṃsakas, for the latter were realists about universals, while the Jainas and the Madhvas rejected, as logically redundant, both universals and inherence. The only difference between Prābhākara and Vaiśeika as regards universals centers on their conceptions of inherence.
Contrasts with Western Metaphysics of Forms and Properties
It should be clear by now that there is no core theory of universals shared by all the Indian philosophers. But one can discern five broad features that distinguish the Indian theories of universals from their Western counterparts:
(1) Even the strongest realist position of the Nyāya-Vaiśeika never took the form of the realism of Plato's theory of ideas. Indian realists about universals were equally realists about the perceptible particulars of the external world. Earthly particulars were never thought to be less real copies of thinkable universals, even by those who believed in universals.
(2) Even if one concedes that the Nyāya universals were closer to Aristotle's universal properties, which are immanent in the worldly particulars, Aristotle could never agree that universals are themselves directly perceived, which is the standard Nyāya position.
(3) The peculiar form that nominalism took in the Indian Buddhist theory of word-meanings as exclusions does not have any parallel in the West. One finds an interestingly different counterpart of the Jaina and Madhva theories of resemblance in Nelson Goodman, but exclusion-nominalism remains a unique contribution of Indian Buddhism.
(4) Most Western realist accounts of universals take colors and such qualities, as well as relations such as "being larger than," as paradigm examples of universal properties. In Indian realist thought the distinction between such particular qualities (guṇa ) and universal properties (jāti ) has been sacrosanct. It is only recently that the idea of particular qualities is gaining ground in Western analytic metaphysics of tropes. Even relations are not treated as genuine universals by any classical Indian realist.
(5) The controversial and complex theory of inherence as a single concrete connector joining not only universals and their instances but also particular qualities to substances and, most puzzlingly, wholes to their parts is totally foreign to the Western realists.
Udayana. Kiranavali. Calcutta, India: Bibliotheca Indica, 1956.
Jayanta Bhāṭṭa. Nyayamanjari. Part 2, Chapter 5. Mysore, India: Oriental Research Institute, 1983.
Dharmakirti. Pramana Varttikam, edited by Swami Dwarikadas Shastri. Varanasi, India: Bauddhabharati, 1994.
Asoka Pandita. "Samanyadusanadikprasarita." In Six Buddhist Nyāya Tracts, edited by Haraprasad Shastri. Calcutta, India: The Asiatic Society, 1989.
Patañjali. Vyakarana-Mahabhasyam. Vols. 1–2, edited by Bala Shastri. Delhi, India: Pratibha Prakashan, 2001.
Bhattacharya, Gopinath. "Some Aspects of the 'Universal.'" In Essays in Analytical Philosophy. Calcutta, India: Sanskrit Pustak Bhandar, 1989.
Chakrabarti, Kisor. "The Nyaya-Vaisesika Theory of Universals." Journal of Indian Philosophy 3 (1975): 363–382.
Dasgupta, Surendranath. The Mahābhāsya of Patañjali: With Annotations, edited by Sibajiban Bhattacharyya. New Delhi, India: Indian Council of Philosophical Research, 1991.
Dravid, Raja Ram. The Problem of Universals in Indian Philosophy. Delhi, India: Motilal Banarsidass, 1972.
Halbfass, Wilhelm. On Being and What There Is: Classical Vaiśesika and the History of Indian Ontology. Albany: SUNY Press, 1992.
Hayes, Richard P. Dignāga on the Interpretation of Signs. Dordrecht, Netherlands: D. Reidel, 1988.
Herzberger, Radhika. Bhartrhari and the Buddhists: An Essay in the Development of Fifth and Sixth Century Indian Thought. Dordrecht, Netherlands: D. Reidel, 1986.
Mukhopadhyaya, P. K. Indian Realism. Calcutta, India: Firma K. L., 1984.
Shastri, Dharmendra Nath. Critique of Indian Realism. Agra, India: Agra University, 1964.
Siderits, Mark. "More Things in Heaven and Earth." Journal of Indian Philosophy 10 (1982): 187–208.
Arindam Chakrabarti (2005)