Gelug Cognition Theory and Understanding of Emptiness

Cognition Theory: Cognition of Commonsense Objects

Last session we looked at several unique points that Tsongkhapa made and contrasted some of them with the non-Gelug positions on them. Since the presentation I gave was quite sophisticated and advanced, I’d like to start this session by going over some of the salient points and explaining them once more, but now in a simpler way.

Tsongkhapa was really a revolutionary reformer who basically reinterpreted many of the teachings, particularly on Prasangika. He also very radically reformed the understanding and explanation of cognition theory, of how we know things. Cognition theory is a fundamental part of Buddhist education; it’s a topic called lorig (blo-rig) in Tibetan, which is usually translated as “ways of knowing.” Tibetans usually spend a year studying this in the monasteries.

According to the Gelug presentation of the Sautrantika system, there are seven different ways of knowing – some are valid, and the others are not valid. There are two valid ways of knowing things, either through what’s called bare cognition (mngon-sum tshad-ma), which is non-conceptual, or through inferential cognition (rjes-dpag tshad-ma), which is conceptual. There are three types of inferential cognition. With inferential cognition we know something obscure by relying on a line of reasoning,  or on some sort of convention, or on a valid source of information. For example, how do we know when our birthday is? We need to ask somebody who was there when we were born, for instance our mother. We couldn’t know it through bare cognition or by reasoning with logic; we know it by relying on a valid source of information. 

Anyway, there was a big discussion about how bare cognition, specifically bare sensory cognition like seeing, hearing and so on, works. There are two interpretations, known as False Aspectarian and True Aspectarian and they differ about what we actually see with the bare cognition of our eyes. This the question. What do we see? For instance, if we look at some object conventionally called an “orange,” the masters before Tsongkhapa and the Sakya, Kagyu and Nyingma masters after him all asserted the False Aspectarian interpretation. Namely, all that we see in any moment is just an orange-colored round shape and we only see the present moment of that. That’s all that we see. 

Only the form or also the color? 

Vaibhashika said we see color and shape separately, but all the other India tenet systems say we do not see them separately but rather we see an orange-colored round shape, which is an arrangement of orange-colored pixels. If we hold an orange in our hand, all that we know through tactile cognition is a certain tactile sensation and only one moment of it at a time. The technical term for what we cognize is “sensibilia,” sensory information. We only know the sense information of one sense with each type of sensory consciousness, although we may simultaneously cognize the information of two senses, such as when both seeing someone and hearing them speak. 

Now, what is an orange? An orange is not an orange-colored round shape. An orange is not just a physical sensation. An orange is not just a smell, and an orange is not just a taste. According to the non-Gelugpas, what we would call a commonsense orange is a conceptual construct, a conceptual synthesis. It is an imputation phenomenon – something that cannot exist or be cognized separately from a basis. The basis is all the sensibilia of the object over a period of time. In addition, a commonsense orange, as an imputation phenomenon that can only be conceptually cognized, is cognized through the concept or category of “an orange.” 

All the sense information regarding this object is changing every moment. What we would call a commonsense orange is something that extends over all the information of all the senses, and it extends over time. But all we can ever know is the information of one sense in one moment at a time. So, you see the problem here. How do we ever know a commonsense orange that extends over time and extends over all the sense information? 

The conceptual mind synthesizes and conceives of a commonsense object. According to this interpretation, we never actually see a commonsense orange, we never taste a commonsense orange. All we see is an orange-colored round shape and all we taste is a specific taste and only one moment of it at a time. That’s very profound. In order to understand the whole discussion of non-conceptual cognition, which is emphasized so much, particularly in Kagyu, Nyingma and Sakya, we have to understand that. Otherwise we get the wrong idea from the discussion in their mahamudra and dzogchen texts about non-conceptual cognition and the importance of that in meditation. I find it a very lovely point and really very profound. 

In other words, all that we ever know through the senses, non-conceptually, is the information of one sense and only one moment of it at a time. That’s all we know. However, a commonsense object is not just an orange-colored round shape. A commonsense orange doesn’t exist for just one moment; it extends over time. We think an orange is sitting on the table over there all day long, but how do we know that? We know that only through the concept of an orange that extends over time, because in any moment all we see is one moment of an orange-colored round shape, and it’s changing every moment, of course. 

In short, according to the non-Gelug False Aspectarian interpretation, a commonsense object is something that is only known conceptually. It is a conceptual synthesis of a sequence of moments of a sight, a taste, a smell and a physical sensation that we can non-conceptually cognize. 

Are Commonsense Objects Seen in Common by Everybody Who Looks?

Do you mean by “commonsense object” that this is seen in the same way by everybody? 

That’s another question and a very tricky one. It depends on whether we are analyzing from a Sautrantika point of view, which asserts the external objective reality of sensory information, or the Chittamatra view that refutes external existence and so refutes external phenomena. If we accept the external objective reality of sensory information, sensibilia, then we would have to say that everyone sees the same orange-colored shape, but of course everyone sees it from a different angle. The concept of an orange, which is equivalent to the conceptual category “orange,” might be shared in common, but what would represent it and appear in conceptual cognition might be individual. I am just speculating about it here. 

If we follow the False Aspectarian Chittamatra view as espoused by the non-Gelug Tibetan traditions, then sensibilia that appear in non-conceptual cognition plus the sensory consciousness and accompanying mental factors that cognize the sensibilia both arise from the same karmic tendency or seed and not from separate natal sources (rdzas). That karmic tendency may be specific to one individual, or it may be a karmic tendency from what is known as “shared karma.” 

Still, the question arises in both Sautrantika and Chittamatra as to whether there is a common locus (gzhi-mthun) of sensibilia – a single set of sensibilia – found the same, but from different angles, in everyone’s non-conceptual visual cognition or is there no common locus for what appears. 

I don’t know the opinion of the various masters from the non-Gelug traditions concerning this point. The False Aspectarian Chittamatra analysis of conceptually cognized commonsense objects is probably the same as what I tentatively postulated regarding the Sautrantika view. In any case, non-Gelug, following the False Aspectarian interpretation of cognition never say that we see a commonsense orange, let alone that we all see the same one. But according to them do we all conceive of the same commonsense orange? This becomes a very profound, difficult question. So, when you ask whether there is an actual orange sitting there and everybody is seeing the same orange, that can become a very deep question.

What Is a Commonsense Object? 

How do you define exactly “commonsense object?” 

A commonsense object is defined as one that extends over the sensibilia – the sense information – of many senses and extends over time. That’s what we ordinarily think of as an orange, or a human being, or anything, isn’t it? These are interesting things to think about. 

In addition, the non-Gelug schools, the earlier schools, were saying that what we non-conceptually cognize with our senses arises from and thus is the result, the effect, of the objective sensory information, if we assert external phenomena in accord with the Sautrantika tenets. That sensory information is the cause, and the seeing of it is the result. The result arises after the cause occurs; they are not simultaneous. What we actually see is a mental semblance or representation, a mental hologram or something like that; we see a mental semblance of an orange-colored round shape. That mental semblance occurs in moment two and the objective, orange-colored round shape occurs in moment one. 

Another question, of course, is do we see just an orange-colored round shape or do we see an orange-colored spherical shape. Do we see depth? Afterall, we don’t see the back side of anything, we see just the front side when we look from the front. So, do we see two-dimensionally, three-dimensionally or, if we add time, four-dimensionally? What do we actually see? 

In any case, the orange-colored round shape of moment one is the natal source that gives rise to a mental aspect (rnam-pa) of itself – a mental semblance or representation, like a mental hologram, of an orange-colored round shape in moment two. While we’re seeing this mental representation in moment two, that original moment one of the orange-colored round shape no longer exists. We never cognize anything directly with our physical senses – our eye-sensors, ear-sensors and so on. Although False Aspectarian Sautrantika asserts external sensibilia, we never see them directly. These mental semblances are opaque: we don’t see anything through them; we just see a mental semblance. We always see, in a sense, something like an afterimage. If we think in terms of the speed of light, that makes perfect sense. We can see the light of distant stars, but that light was emitted millions and billions of years ago. 

If we pursue the logic of it, it starts to get really very freaky. Because also, if we may bring in a little bit of Western explanation here – the mental semblance we see is first produced on the curved surface of the eye sensor, the retina, and despite the curve, the surface is two-dimensional. So from the Western scientific view as well, the perception of three-dimensional depth is a mental construct.  

I thought that the eyes are put in a certain angle, so that’s why you see three-dimensionally. 

Well, yes, but the sight of depth is still a mental construct. One eye sees one thing; the other eye sees another thing. How are they put together? It’s through a mental construct. Time as well is a mental construct, because all we ever know is one moment at a time. If we pursue this, it starts to affect our understanding of the very nature of reality. This can lead to the view, that some Western theorists espouse, that the whole universe is actually a hologram. 

If we pursue the Western analysis even more, the way that our eyes see things is upside down. That’s even more evidence that our commonsense world is a mental construct. It’s certainly the case with language, even. Think about language. All that we hear is one syllable at a time. It’s certainly a mental construct that puts together those moments of consonants and vowels into a word, let alone into a sentence that has meaning. All of that’s a mental construct. 

The Gelug True Aspectarian Interpretation of the Cognition of Commonsense Objects  

Tsongkhapa took the True Aspectarian approach and said to the False Aspectarians, “What you say makes sense, but it’s a bit too much.” So Tsongkhapa asserted that when we see one moment of an orange-colored round shape, we have to say that simultaneously we also see the commonsense object. We also see the commonsense orange. Even though that commonsense orange is a synthesis of many moments of the sensibilia cognizable through various senses, we have to say that we see a commonsense orange; it’s not that an orange can just be known conceptually. We can’t eat a concept. 

In other words, what Tsongkhapa is saying is that we have to make a distinction within syntheses. It doesn’t make sense to say, like the earlier schools do, that all syntheses are conceptual constructs. Some syntheses can be cognized non-conceptually. If this were not the case,  then there is a big problem regarding a whole and its parts. A whole is a synthesis of parts. So, do we only see parts and never see a whole? In order to see an orange, do we need to see the backside when we look from the front? Is the defining characteristic mark of the whole found in each of the parts or is it a separate defining characteristic mark not found in any of the parts? How do we cognize a whole commonsense object? This is a big debate.  

According to the True Aspectarian position as explained by Gelugpa, there are some syntheses that we can actually see and hear. These refer to whole commonsense objects that extend over all their sensibilia and over time. We have to agree that a whole commonsense object is an imputation phenomenon – something that can only exist on the basis of parts and can only be cognized simultaneously with an appearance of at least some of its parts. Imputation phenomena like these are syntheses. The question is: Are such syntheses merely conceptual constructs or can some be cognized non-conceptually, Tsongkhapa says, “Some types of syntheses you can definitely cognize non-conceptually, and some you can’t. 

What would be an example of a conceptual synthesis, a purely conceptual construct, that we can only cognize conceptually? An example is a category, which is an imputation phenomenon on the basis of individual items that fit into it. A category is a conceptual synthesis based on various items sharing common defining characteristic features. For instance, we see this commonsense object in front of us. But to cognize it as a table, “This is a table,” we can only cognize it as such conceptually. When we cognize it conceptually as a table, a mental hologram of a table appears and we cognize it through the category “table,” designated with the word “table.” In this conceptual way, when we see this object in front of me, that object near you and that object at the back of the room, we can cognize them all as tables and call them all “tables,” although they’re very different looking objects. 

The question then is, “Well, do we see a table?” The True Aspectarian Sautrantika view, which asserts objective reality, says that there are not only whole objects, known as “collection syntheses” (tshogs-spyi) that can be cognized non-conceptually on the basis of parts, there are also kind syntheses (rigs-spyi) that can also be cognized non-conceptually on the basis of individual items that are all the same kind of object, like all are objectively tables or oranges. We can see an object and what we are seeing is objectively a table, but we can only know it is a table conceptually through the category “table.” Chittamatra and Madhyamaka, of course, refute that this object’s being a table is established independently from its own side, but from the Sautrantika point of view, this object is actually, objectively a table and not an orange or a dog. 

Tsongkhapa adds in accord with the True Aspectarian interpretation that we when we see a commonsense object through a mental semblance, a mental hologram, that mental hologram is transparent. Otherwise, we get the problematic issue of a time lapse in sensory cognition that the False Aspectarian interpretation asserts with its assertion of opaque mental holograms arising the moment after the occurrence of the sensibilia that the hologram represents. This is why we have to use the words “direct” in the case of the Gelug True Aspectarian interpretation of sensory cognition and “indirect” in the case of the non-Gelug False Aspectarian interpretation. To avoid confusion, we need to reserve these terms “direct” and “indirect” for only this distinction and use “explicit” and “implicit” for another distinction in ways of knowing. After all, there are different technical terms in Tibetan for each of these. 

In summary, Gelugpa asserts that we see commonsense objects directly through transparent mental semblances of them, whereas the non-Gelug Tibetan traditions assert that we see only one moment of the sensibilia of the sense of sight, but not directly. We see it only indirectly because what we actually see is an opaque mental semblance of the sensibilia since, with sensory non-conceptual cognition, there’s a one moment lag in time between the occurrence of the sensibilia and the occurrence of the sensory non-conceptual cognition of it. In addition, only subsequent to the sensory non-conceptual cognition do we cognize a commonsense object and that is only through conceptual cognition. There are some Buddhist masters, for instance Shakya Chogden from the Sakya tradition, who say that we don’t even see colored shapes; all we see are colored pixels. Even colored shapes are conceptual constructs, which is probably even closer to our Western analysis, if we really pursue it. 

I went into this a little bit in detail because I know that many of you haven’t studied Buddhist cognition theory before. Actually, I find it absolutely fascinating, and it really, really gets you thinking about the whole discussion of reality. The discussion of reality is very much connected with the discussion of cognition. What do we actually know? What do we cognize? All we know is what we cognize, either non-conceptually or conceptually. Are what we cognize in either of these two ways actually real? What establishes, what proves, that something exists and is real? From the Gelug Prasangika point of view, the only thing that establishes or proves that something exists as a commonsense object is that there are words and concepts conventionally agreed upon for them and that cognition of objects as the referent objects of these words and concepts can be validated by valid cognition of their superficial and deepest truths. We can’t say anything else. 

Concepts, Mental Labels, and Categories

A concept (rtog-pa), by the way, as explained yesterday, is a conceptual cognition of an individual item (bye-brag) mixed with a category (spyi) in which it is conceptually fit. A category and a mental label (btags) are synonyms. Often, however, in oral explanations especially in Western languages, “concept,” “mental label” and “category” are used interchangeably. 

This was also discussed in art. If you see an abstract painting, then you usually try to see some figures, and you want to conceptualize these abstract figures. And you will make a form – maybe it’s this shape or another shape. So, there it was also discussed. What do you see? You usually try to see a form. You want a form, and you don’t get it. 

Well, this is a sign of our grasping for things to exist as concrete objects. Yes, this occurs when viewing abstract painting or pointillism, where they painted with pixels, That’s a good example. There, the shapes and objects we think that we see are really conceptual constructs. 

Distinguishing That an Object Is “This” and “Not That”

One more point that is good to mention in terms of cognition theory, which follows from what we were discussing, is that according to non-Gelugpa, the earlier theory, determination that something is this and not that occurs only with conceptual cognition, in terms of conventional, commonsense objects. It’s this object as opposed to that object, or this table as opposed to that table, or whatever. When we non-conceptually cognize a colored shape, for instance, we cannot distinguish it as the colored shape of an orange for instance. According to non-Gelug, then, distinguishing is only a conceptual occurrence, whereas Tsongkhapa said, no. Since non-conceptual cognition cognizes commonsense objects, we distinguish and determine that what we cognize is a specific commonsense object with sensory non-conceptual cognition of it. We see commonsense objects and simultaneously determine them to be this and not that

That distinguishing occurs only conceptually according to non-Gelug has many ramifications in terms of their mahamudra and dzogchen meditation, which instructs us to go beyond distinguishing between this and that. This instruction is based on their assertions about cognition. Since distinguishing between “this” and “that” is exclusively a conceptual occurrence, then to meditate non-conceptually, we need to go beyond making such distinctions. 

As I say, in order to really understand properly things like mahamudra and dzogchen meditation in the non-Gelug schools, we really have to know the assertions of their cognition theory, especially when they’re talking about becoming non-conceptual. Tsongkhapa changed the understanding of non-conceptuality very much, so the whole discussion of non-conceptual and conceptual cognition is very different in Gelug. 

There are many other points about cognition theory, but probably this is enough because there are many other topics as well. So, let’s get back to what we started yesterday, which is Tsongkhapa’s revision of Prasangika. In connection with cognition theory, Tsongkhapa also asserted that Prasangika explains several points differently from Sautrantika and Chittamatra. For example, bare cognition is exclusively non-conceptual in Sautrantika and Chittamatra. It is “bare” of categories and concepts. Bare cognition is redefined in Prasangika as straightforward cognition, since it does not rely on a line of reasoning, and therefore may be either non-conceptual or conceptual. None of the non-Gelug traditions made any distinction in terms of a Prasangika interpretation of cognition. Remember, according to the non-Gelug, Prasangika does not make any positive assertions about anything. 

Cognition of Voidness According to Prasangika 

This whole discussion of conceptual and non-conceptual cognition affects very much the way that Tsongkhapa presents the Prasangika view of voidness and voidness meditation. Remember, the central issue that affects so much of Buddhist meditation and philosophy is how do we go from a conceptual to a non-conceptual cognition of voidness. Obviously, this is going to be very much affected by our understanding of what it means to be non-conceptual. What actually is conceptual cognition? What actually is non-conceptual cognition? We need to fit everything  together here. 

The non-Gelug schools both before and after Tsongkhapa – Kagyu, Nyingma and Sakya – assert that the voidness that we validly cognize conceptually and validly cognize non-conceptually are different voidnesses; they’re not the same voidness. The voidness that we cognize conceptually is just a mental construct of voidness. The voidness that we cognize non-conceptually is a voidness beyond words and concepts. Tsongkhapa said they’re the same voidness whether we cognize it conceptually or non-conceptually. 

The non-Gelug position is that since all conventional, commonsense objects are conceptual constructs, they are only cognizable conceptually. These conventional, commonsense objects that appear in conceptual cognition appear to have truly established existence, which means true, unimputed existence – existence established independently of imputation by conceptual cognition. Unlike the Gelug assertion that it is only in terms of mental labeling, which is always conceptual, that the conventional existence of commonsense objects can be established, non-Gelug asserts that conceptual mental labeling actually synthesizes and creates conventional, commonsense objects. This is a radical difference. 

The non-Gelug traditions assert, then, that not only is truly established existence a conceptual construct, so too the voidness of truly established existence is merely a conceptual construct. In other words, since truly established existence is a conceptual construct that does not refer to anything that is even conventionally findable, then both the affirmation and the negation of truly established existence are equally also conceptual constructs and, as conceptual constructs, both appear to be truly existent and therefore both are false. Conceptual cognition of the affirmation and the negation of truly established existence, then, is like the conceptual cognition of the birth and the death of the child of a barren woman – a woman who is sterile. 

So whether we cognize the existence of truly established existence or the non-existence of truly established existence, both are merely conceptual cognitions and ultimately false. To attain cognition of the deepest truth, ultimate reality, we have to go beyond conceptual cognition. We need to go beyond both the sensory non-conceptual cognition of single moments of the sensibilia of one sense, even though they do not appear to be truly established commonsense objects, and we need to go beyond the conceptual cognition of commonsense objects, including the concept of voidness. We need to gain a cognition of voidness that is beyond words and concepts. This is the object cognized without any concepts by an arya’s deep awareness. This voidness beyond words and concepts, then, is not the voidness of truly established existence that is discussed in the texts and realized through logic. 

Tsongkhapa says, “No, no. It’s not like that.” The voidness of truly established existence is something that can be known non-conceptually. It’s the same voidness that’s known conceptually and non-conceptually. See, it’s parallel to this discussion of commonsense objects. Can we actually know commonsense objects non-conceptually? Tsongkhapa says, “Yes. We can know them both conceptually and non-conceptually and it’s the same thing with voidness. The same voidness can be known both conceptually and non-conceptually. The discussion about how to go from a conceptual to a non-conceptual cognition of voidness deals only with the mind that cognizes voidness and how it cognizes it – either through a concept or not through a concept – in other words, through the filter of a category, the category “voidness,” or not through such a filter. Going from a conceptual to a non-conceptual cognition of voidness does not entail a change of the object of cognition, voidness itself. Voidness is voidness. The only thing that’s different is our way of cognizing it. 

The Validity of Conceptual Cognition

This is a major, major difference. Now, we’re getting to the real radical things that Tsongkhapa was saying. Tsongkhapa asserted that although conceptual cognition entailed grasping for truly established existence, nevertheless conceptual cognition can still be valid, as in the case of valid inferential cognition. Bodhichitta, for example, has to be conceptual all the way up to enlightenment since it is focused on enlightenment and only a Buddha can non-conceptually cognize enlightenment. Conceptual cognition, then, is not an obscuration preventing either liberation or the omniscience of Buddhahood. It is not the main thing we need to work on ridding ourselves of. Sure, we ultimately need to rid our minds of it, like abandoning the boat once we have crossed a stream, since Buddhas do not have conceptual cognition. But that doesn’t mean that the boat of conceptual logic and conceptual bodhichitta is completely false and unreliable. 

True, the non-Gelug traditions do not dismiss logic and conceptual thought as being useful methods, but the attitude toward them is very different from the attitude Gelugpa has toward them. Traditionally, the non-Gelugpas, especially Nyingma and Kagyu, put a big emphasis on non-conceptual meditation and downplay logic and debate. Gelugpa puts the emphasis on the use of logic and debate to clear away any inaccuracies or any indecisiveness about our understanding as a pathway for ensuring successful meditation. 

Grasping for True Existence

To get back to our discussion, grasping for true existence entails two aspects: giving rise to an appearance of true existence and grasping for that appearance to correspond to the actual way in which something exists. The non-Gelugpa traditions assert that this distinction of two aspects of grasping for true existence pertains just to conventional, commonsense objects. Since the appearance of conventional, commonsense objects only arises in conceptual cognition, both aspects of grasping for true existence occur only in conceptual cognition. When an appearance of a conventional, commonsense object arises, it can only be arising in a conceptual cognition and at that time it appears to be truly existent, and the mind grasps for that appearance to correspond to reality. Since conventional, commonsense objects do not arise in sensory non-conceptual cognition, there is no appearance of true existence and no grasping for such an appearance to correspond to reality in sensory non-conceptual cognition.

Sensory non-conceptual cognition does not know how things actually exist, but at least it doesn’t have the fault of making appearances of commonsense objects that truly appear to exist as fitting into the categories “this” and “not that.” So, to stop conceptual thought and get to merely a non-conceptual sensory state is a really important step in the non-Gelugpa systems toward attaining non-conceptual cognition of voidness beyond words and concepts. This is because during sensory non-conceptual cognition at least our minds don’t make appearances of true existence, because with sensory non-conceptual cognition our minds do not conceptually synthesize and give rise to appearances of conventional, commonsense objects. 

Tsongkhapa said, “Both conceptual and non-conceptual cognition give rise to an appearance of true existence because we actually cognize and distinguish conventional, commonsense objects as this or that even with sensory non-conceptual cognition. The Jetsunpa textbooks within Gelug say that although our sensory non-conceptual cognition does not grasp for that appearance of true existence to correspond to the way in which things actually exist, nevertheless that non-conceptual sensory cognition is accompanied by a subliminal conceptual cognition that does grasp for it to correspond. The Panchen textbooks say that the grasping for the appearance of true existence to correspond to reality continues during sensory non-conceptual cognition as merely the habit for grasping for true existence. Whether we follow the Jetsunpa assertion or the Panchen one, in either case although stopping the thought processes of conceptual cognition may be helpful in concentration meditation for quieting the mind of mental wandering, attaining sensory non-conceptual cognition will not help in ridding our minds of grasping for true existence.  

From a Madhyamaka point of view, both the Gelug and non-Gelug traditions assert that truly established existence does not correspond to how things actually exist. According to Gelug, truly established existence is defined in Madhyamaka as the existence of something being established as true independently of it being merely the referent object of the mental label and word for it. In other words, the belief that the findable existence of something can be established or proven out of the context of its being merely the referent object of a mental label or word for it. Although this findable existence appears to be established as if it were true, nevertheless it’s not true. True existence is false according to Madhyamaka. 

The Definition of Truly Established Existence in the Various Tenet Systems

If you talk about true existence, then you mean, according to Nagarjuna, that it is not findable and it’s just only a construction of the mind and that it exists not out of itself? 

Unfortunately, truly established existence is defined differently in the different Indian tenet systems, so that doesn’t make it easier at all. Also, the use of the word “existence” in the discussion of this topic of voidness in Western languages can also be misleading. The Sanskrit and Tibetan terms that are translated as true existence do not contain the word “existence.” The Sanskrit word that is used is siddha, Tibetan drub (grub), which means “established” or “proven.” So, we need to understand that the whole voidness discussion concerns the refutation of impossible ways of establishing the existence of phenomena and assertions of the valid ways to establish their existence. 

What establishes or proves that something exists –that’s the issue. That’s what the whole discussion is about. Because we can only know something in terms of the mind, how do we know that what we cognize corresponds to reality. What establishes or proves the existence of what we cognize? Is its existence established from something on its own side? Or can we establish that it exists only in terms of the mind that cognizes it? 

Maybe we can understand the discussion a little more easily if instead of the word “true” we substitute the word “real.” What establishes that something is real? Let’s stick to just the Gelug assertions. Gelugpas say that the Sautrantika tenet system asserts that if something produces an effect, it’s real. If it doesn’t produce an effect, it’s not real. Or if we can see something it’s real. If we only can think it, it’s not real; it’s only our imagination. This is the Sautrantika explanations and, for them, phenomena that have truly established existence are objective entities and real. 

If I may simplify their assertions like that, then I think we can get the idea. How do we know anything is real? The wall’s real because we bang into it if we try to walk through it. It produces an effect. Or what’s the difference between a person that is actually here in this room and the hallucination of someone being here or imagining that some person is here. If we can actually see the person and other people can see this person too, what we all see corresponds to reality and is real. If no one else can see this person as being here, then what we see isn’t real. It’s a hallucination. We also think like that in the West. 

Vaibhashika asserts that both nonstatic and static phenomena have truly established existence, because even static, unchanging objects give rise to the cognition of them. Sautrantika, on the other hand, asserts that only nonstatic phenomena – phenomena that change in each moment – have truly established existence and are objectively real, whereas static phenomena such as conceptual categories are metaphysical entities and ultimately not objectively real. They’re conceptual, they’re imaginary. 

Chittamatra redefines truly established existence and defines it in several different ways. But to make it simpler, let’s just say that for Chittamatra, truly established existence is existence established from something’s own side, independently from being merely the referent object of a mental label or word for it,  Thus, nonstatic phenomena and voidness are truly established and are real. So, it’s not the fact that there are names for them or mental labels or concepts for them that prove that they exist and are real. Of course, there are names for them. That doesn’t prove anything, Chittamatra says. After all, there are names for imaginary, totally conceptional phenomena too, like turtle hair and externally established existence. But imaginary objects do not have findable defining characteristic marks on their own side that establish their existence as real. Only nonstatic phenomena and voidnesses have such findable characteristic marks and so only they have truly established existence and are real. The fact that there are names, mental labels and concepts for imaginary objects does not establish them, then, as real. 

Now, we get to Svatantrika and, according to the Gelug explanation, Svatantrika says that nothing has truly and unimputedly established existence. The conventional existence of anything can only be established in terms of it being the referent object (btags-chos) of the name and concept for it, however – and that’s a big “however” – its existence cannot be established merely in terms of it being the referent object of the name and concept for it. There must also be a defining characteristic mark (mtshan-nyid) findable on the side of both the basis for labeling (gdags-gzhi) and the referent object of the mental label of a category that allows the referent object to be correctly fit into the mental label of a category and correctly designated with the name or word ascribed to that category. Being the referent object of a word or mental label in conjunction with having the appropriate findable defining characteristic mark on its own side and on the side of its basis for labeling is what establishes the existence of something. 

Furthermore, Svatantrika asserts that conventional phenomena have self-established existence (rang-bzhin-gyis grub-pa), usually translated as “inherent existence.” This is existence established by a findable referent “thing” (btags-don), which is a focal support (dmigs-rten) for the referent object of the word or mental label for it, like a prop holding up a piece of scenery on a stage. The referent “thing” is what actually has the findable defining characteristic mark of the referent object on its own side. It is because an aspect of the referent “thing” appears as the referent object in a conceptual cognition that establishes that the referent “thing” conventionally exists and, in this sense, is real. Ultimately, however, only the deepest truth, the voidness of the truly and unimputedly established existence of the referent “thing,” arises in the non-conceptual total absorption of an arya.

Then, Prasangika says that although all phenomena have defining characteristic marks that allow us to distinguish one thing from another, there are no such things as findable characteristic marks on the side of phenomena that even in conjunction with the phenomena being the referent objects of the words and mental labels for them establish their existence. And even though an appearance of self-established existence arises, the fact that it appears does not establish that self-established existence conventionally exists. Self-established existence and referent “things” as the focal support of phenomena disintegrate and perish when subjected to the force of logic. The only thing that establishes the existence of something is that it is merely the referent object of a word or concept for it and that cognition of it as being this referent object is not contradicted by valid cognition of either its conventional or deepest essential nature. There is nothing findable on the side of phenomena that has the power – either alone or in conjunction with being the referent object of a word or concept – to establish the existence of anything. 

The Svatantrikas, then, were saying that valid and invalid labeling is determined from the side of an object by its findable defining characteristic mark, while Prasangika is saying that valid or invalid labeling is determined from various criteria from the side of the mind and none of these criteria is findable, of course. 

Mental Labeling

As we can see, Tsongkhapa’s formulation of Prasangika is very different from what the other Tibetan traditions had said before. Sakya, Nyingma and Kagyu, following a False Aspectarian interpretation of cognition, assert that the Prasangika and Svatantrika methods both lead to the same cognition of voidness beyond words and concepts. Prasangika makes no assertions of its own. Svatantrika – referring to what Gelugpas call Yogachara-Svatantrika – makes all the assertions. According to the False Aspectarian interpretation, conventional truth – referring to conventional, commonsense objects – is cognizable only by conceptual cognition. Existing merely as the conceptually synthesized referent objects of the words and concepts for them, conventional objects are not found by analysis with logic, even though logic is conceptual. Conventional, commonsense objects are all false. Deepest truth is voidness beyond words and concepts. It is what is found by the non-conceptual analysis or discernment of an arya’s deep awareness (ye-shes) and, in this sense, is truly existent in contrast with commonsense objects. 

In short, according to non-Gelug, conventional objects are merely the referent objects of words and concepts, they can only be cognized conceptually, and all conceptual cognitions are false. Deepest truth voidness is beyond being merely the referent object of words and concepts. It can only be cognized by an arya’s deep awareness and not by either conceptual cognition or an ordinary being’s non-conceptual cognition, and it is true. 

Gelug does not assert that conventional objects exist merely as the referent objects of the words and concepts for them. Gelug asserts that the existence of conventional objects can be established merely by their being the referent objects of the words and concepts for them. That existence established merely in terms of mental labeling can be validated by accurate cognition of both the objects’ superficial and deepest essential natures – and that refers to both accurate non-conceptual cognition and accurate conceptual cognition of the two essential natures that all validly knowable objects have. 

When we talk about mental labeling in Gelug Prasangika, the discussion concerns what establishes that something conventionally exists. In the non-Gelug systems, the discussion of mental labeling concerns what creates the conventional existence of something, and that conventional existence is always false – conventional objects don’t really exist. This is a huge difference. Gelug argues that merely the fact that there is a name, a mental label or a concept for something doesn’t create it, even in the case of a hallucination. Also, existence established merely in terms of mental labeling doesn’t mean that for something to exist, we have to actively be labeling it or conceiving of it now. It never means that. Whether anybody names something or not doesn’t matter at all, but how do we know that something exists? What establishes that it exists? Just the fact that there’s a name for it? No. How else could we know that it exists? From the side of the object? No. If what established the existence of something were just on the side of the object, how would we know that independently of a mind or a measuring device? It’s absurd to say that it’s on the side of the object. 

Tsongkhapa uses the example, which is a very good example, of how would we know that the universe exists in the eons before sentient beings inhabited this universe. Well, nobody was around to mentally label it. The absurd conclusion would follow that if there was nobody around to actively mentally label the universe, the universe wouldn’t exist; it didn’t exist before there were beings in, unless we interpolate the concept of a creator god onto Buddha and say, “Well, Buddha mentally labeled it.” But that’s absurd. 

So, what establishes that something exists is that it can be mentally labeled; there is a conventionally agreed-upon name for it and that name refers to something, but nobody has to be around labeling it. The non-Gelug systems, you remember, do not accept external existence. Therefore, they would probably reject any debate as impossible regarding the question of whether the universe existed externally before any sentient beings were around to mentally label it. The Gelugpas and non-Gelugpas would not have a topic of debate, external existence, that was mutually acceptable. It would be like a theist wanting to debate with an atheist whether God the creator was male, female or beyond any gender.  

It is a common misconception about Gelug Prasangika that things are created by mental labeling, since that leads to the misconception that it’s only when we mentally label something that it exists. That we really have to understand in order to understand Gelug Prasangika correctly. Otherwise, we fall to a totally solipsistic view that everything just exists in our minds and so we imagine, “I am the center of the universe and I create everything by my concepts.” That’s not even the Chittamatra Mind-Only view and certainly not the non-Gelug Madhyamaka view either.

The Greatness of Tsongkhapa

In terms of our topic, Tsongkhapa’s explanation of Prasangika was totally revolutionary. Nobody before him explained like he did. Tsongkhapa proved everything he asserted on the basis of what Nagarjuna said. He showed that this is actually what Nagarjuna was saying and that the other people earlier didn’t really interpret Nagarjuna’s words correctly; they interpreted just one level of their meaning, but not their ultimate,  deepest level. 

Tsongkhapa was really a genius, like the Einstein of his time. But look at what he did in order to gain the insights that he had. He did 3.5 million prostrations and 1.8 million mandala offerings in order to build up sufficient positive force to break through the old way of thinking and to get to this new understanding. In the end, Manjushri had told him, “To understand correctly, look again at Buddhapalita.” Buddhapalita was an Indian master who wrote an extensive commentary, mostly called by his own name, Buddhapalita, to Nagarjuna’s root text. “Now, you’ve built up enough positive force to understand really what he was saying.” That’s how Tsongkhapa’s own non-conceptual realization of voidness came about. 

I find how he actually achieved his attainments through building up a vast store of positive force or potential very inspiring. How did he become such a remarkable being? He was extraordinary. He was a revolutionary. Where did it come from? It wasn’t just, “I’m going to be a rebel and tear down what other people think so I can express my individuality.” It wasn’t that. He built up the causes and, when the causes are complete, then results automatically arise through dependent arising.