On Nāgārjuna’s Ontological and Semantic Paradox

Koji Tanaka ∗

School of Philosophy Research School of Social Sciences Australian National University Koji.Tanaka@anu.edu.au

Abstract

Garfield and Priest have rationally reconstructed Nāgārjuna’s doctrine of emptiness as an endorsement of contradictory nature of reality. According to them, Nāgārjuna can be seen to be arguing that the way in which things exist in reality and what we can truly say about them must be contradictory. What would be the reasons for thinking that Nāgārjuna would accept their radical interpretation? In raising this question, I am not concerned with how their interpretation coheres with Nāgārjuna’s texts. Rather, I am concerned with the internal coherence (or consistency) within their interpretation. I will reject their interpretation by identifying the incoherence within the resources that Garfield and Priest themselves find in Nāgārjuna.

Introduction

In one of his key texts, Mūlamadhyamakakārikā (MMK), Nāgārjuna famously sets out to refute the ontology of essence.i He presents numerous arguments to show that things don’t exist essentially — that is, that things are empty of essence or inherent existence. The doctrine of emptiness has been variously understood by traditional and contemporary commentators. Most radical is the recent interpretation presented by Garfield and Priest (2003). They have rationally reconstructed Nāgārjuna’s doctrine of emptiness as an endorsement of contradictory nature of reality. According to them, Nāgārjuna can be seen to be arguing that the way in which things exist in reality and what we can truly say about them must be contradictory.ii In order to have a grip on their interpretation, compare it with another interpretation according to which Nāgārjuna was a mystic. There are several versions of mystical interpretations and not all of them may come down to the same. According to the mystical interpretation that Garfield and Priest, inter alia, argue against, however, if reality is empty, as Nāgārjuna claims it to be, then there is nothing that can coherently or rationally be said about it. Whatever expressions about reality we can assert, they end up being contradictory. Hence, so this mystical interpretation goes, we should be silent and say nothing about reality. Under this interpretation, emptiness is the reason for Nāgārjuna’s mysticism.iii Garfield and Priest argue against this mystical interpretation by rejecting the inference from the contradictory nature of our expressions about reality to the necessity of being silent about reality. According to them,

not all contradictory expressions about reality are to be rejected and N āgārjuna embraces the contradictory nature of reality rather than reject it by remaining silent. There are a number of important disagreements between Garfield and Priest, on the one hand, and the mystics, on the other. One disagreement is about the principle of reasoning that they think can be attributed to Nāgārjuna. The mystics (according to Garfield and Priest) subscribe to the classical principle of logic that all contradictory expressions are meaningless and trivial. Armed with paraconsistent logic which does not automatically reject contradictory expressions as meaningless and trivial,iv Garfield and Priest reject exactly this principle. Their disagreement is not just about the classical principle of logic, however; there is a deeper disagreement between them about the way Nāgārjuna views reality. According to Garfield and Priest, Nāgārjuna takes reality to be contradictory (or inconsistent) whereas, according to the mystics, he does not hold reality as contradictory even though our expressions of it are (necessarily) contradictory. For Garfield and Priest, emptiness is Nāgārjuna’s analysis of the contradictory (or inconsistent) way in which things exist; whereas for the mystics, it is just a means for silent contemplation on consistent, ineffable reality. So, according to Garfield and Priest, Nāgārjuna presents an ontological thesis that the way things exist is contradictory. Under their interpretation, Nāgārjuna presents an ontological paradox by means of his arguments for emptiness. Garfield and Priest interpret Nāgārjuna to argue that, since to be empty is to have emptiness, emptiness is the nature of all things. This, however, seems to be a claim about fundamental ontology, the very possibility of which is what Nāgārjuna sets out to refute by means of emptiness. In this way, Nāgārjuna faces an ontological paradox, so Garfield and Priest argue. This is a radical interpretation given that neither Nāgārjuna nor the Madhyamaka tradition (the tradition that arose as the result of Nāgārjuna) had ever been seen as accepting contradictions. In fact, Nāgārjuna himself seems to reject the appearance of the contradictory nature of his doctrine. First, his argument proceeds by rejecting all the catuṣkoṭis and Nāgārjuna’s rejection of the third koṭis implies that he is prepared to reject all contradictions.v Second, in the Vigrahavyāvaryanī, he seems to be aware that someone might think of his doctrine as contradictory. The adversary charges that Nāgārjuna’s thesis of emptiness is either contradictory or ineffectual. In response to this, Nāgārjuna rejects both hones of the dilemma by claiming that he has no thesis of his own.vi Given that he rejects any suggestion of the contradictory nature of the thesis of emptiness, it is hard to imagine Nāgārjuna as accepting any contradictions. Third, Mādhyamikas have developed various mechanisms to defuse the appearance of contradictions. Tsong kha pa, for example, had the method of parametarising statements which have the effect of preventing A and not-A to be conjoined. This prevents contradictions of the form A & not-A from arising within Madhyamaka.vii So, what would be the reasons for thinking that Nāgārjuna would accept their radical interpretation? Even if we agreed with Garfield and Priest that Nāgārjuna wasn’t a mystic (though a refutation of the mystical interpretation would require a careful examination), viii it would be hard to agree with their radical interpretation. Garfield and Priest acknowledge that their interpretation deviates from the traditional interpretations. They also acknowledge that Nāgārjuna didn’t explicitly endorse any contradictions (even though they claim that Nāgārjuna would endorse the contradictory nature of reality). Yet, they claim that, once we follow through his doctrine of emptiness to its logical end, we should think of him as holding the view that reality is ultimately contradictory. Given that they are not fazed by the traditional attempts to defuse any appearance of contradictions, we can’t rely on the tradition to critique Garfield and Priest’s radical interpretation.ix We have to identify incoherence within the resources that they themselves find in Nāgārjuna if we are to reject their interpretation. This is what I will do in the rest of this paper. Thus, I am not concerned with how their interpretation coheres with

Nāgārjuna’s texts. Rather, I am concerned with the internal coherence (or consistency) within their interpretation.

Semantic and Ontological Paradox

Garfield and Priest argue for their claim that Nāgārjuna presents reality as contradictory based on their analysis of the doctrine of emptiness as presented in MMK. Nāgārjuna explains the doctrine in terms of two realities – ultimate and conventional realities – and Garfield and Priest follow his explanation. As we will see, they focus on ultimate reality in order to attribute paradoxes to Nāgārjuna. The doctrine of emptiness is that nothing exists with svabhāva. Nāgārjuna presents numerous arguments to show this.x Given that svabhāva (understood as essential property or essence) is what ultimately gives an object its numerical identity, the doctrine of emptiness means that nothing ultimately exists.xi A consequence of this, as Garfield and Priest acknowledge, is that ‘there is no ultimate reality’ (p. 10). This doesn’t mean that Nāgārjuna was a nihilist (Cf., Garfield (2014) and Garfield and Siderits (2013)); rather, it means that there is no ultimate way in which things exist, i.e., if something exists, it doesn’t exist ultimately (with svabhāva). Corresponding to ultimate and conventional realities, ‘there are two truths: conventional truths, the truth about conventional reality; and ultimate truth, the truth about the ultimate reality – qua ultimate reality’ (Garfield and Priest (2003): 4). But, given that ultimate truths are about ultimate reality and that there is no ultimate reality, ‘[t]here are, therefore, no ultimate truths’ (p. 10). Garfield and Priest argue, however, that `[t]here are ultimate truths’ (p. 11) for Nāgārjuna. They claim, in fact, that Mūlamadhyamakakārikā is full of them. For example, in MMK XXIV:19, he says

Something that is not dependently arisen Such a thing does not exist. Therefore a non-empty thing Does not exist. (Quoted in Garfield and Priest (2003): 11)

Garfield and Priest argue that ‘he is telling us about the nature of ultimate reality’, and in doing so, he is asserting an ultimate truth because it is important for us to grasp it (p. 11). Accordingly,

There are, therefore, ultimate truths. Indeed, that there is no ultimate reality is itself a truth about ultimate reality, and is therefore an ultimate truth! (p. 11)

Thus, according to Garfield and Priest, there are and aren’t ultimate truths for Nāgārjuna. They claim that Nāgārjuna ends up in a paradox of expressibility: he expresses ultimate truths, even though they can’t be expressed since there are no ultimate truths that could be expressed. Putting aside the question whether or not Nāgārjuna is expressing ultimate truths for now, the inference Garfield and Priest draw on behave of Nāgrājuna is unsound. They infer that there are ultimate truths from the premise that ‘that there is no ultimate reality is itself a truth about ultimate reality’ (ibid.). If there are truths about ultimate reality, then there are ultimate truths. Hence, the inference is valid. But before making this inference, they claim that there is no ultimate reality for Nāgārjuna. If this is correct, then there is no truth that corresponds to ultimate reality. So the premise of that inference ⎯ that there is an ultimate truth that there is no ultimate truth ⎯ doesn’t obtain for Nāgārjuna. The inference Garfield and Priest draw must, thus, be unsound for Nāgārjuna. Hence Garfield and Priest can’t use that inference to show that there are ultimate truths for Nāgārjuna.

If Garfield and Priest nevertheless insist that there are ultimate truths for Nāgārjuna, then they would have to be able to explain why he would contradict himself. If, for example, Whitehead and Russell repeatedly asserted that 1 + 1 ≠ 2 while proving that 1 + 1 = 2 in their system, then they would just be contradicting themselves. Importantly, there is nothing profound about the contradiction. What makes Nāgārjuna any different? Garfield and Priest respond to several attempts to save him from inconsistency and reject all of them. It is one thing to reject the attempts to make Nāgārjuna’s thought consistent; it is another to directly explain why Nāgārjuna is contradicting himself for a good reason, however. Can they explain the difference between Nāgārjuna and our imaginary Whitehead and Russell? The main explanation Garfield and Priest provide seems to be that the doctrine of emptiness entails a contradictory (or inconsistent) ontology. As we saw above, Nāgārjuna argues that everything is empty of svabhāva. So, ‘all things lack any ultimate nature’ (p. 14). But, ‘this is a characterisation of what things are like from the ultimate perspective’ (p. 14). They then go on to argue thus:

Thus, ultimately, things are empty. But emptiness is, by definition, the lack of any essence or ultimate nature. Nature, or essence, is just what empty things are empty of. Hence, ultimately, things must lack emptiness. To be ultimately empty is, ultimately, to lack emptiness. In other words, emptiness is the nature of all things; by virtue of this they have no nature, not even emptiness. (pp. 14-15)

According to Garfield and Priest, thus, everything is empty of nature. But, because that is how everything ultimately exists, emptiness is its ultimate nature. So emptiness is not ultimately empty. But, because everything is empty of ultimate nature, emptiness must also be ultimately empty. Thus, ultimate reality is both empty and not empty. Hence, so Garfield and Priest argue, Nāgārjuna would have to be committed to a contradiction. Garfield and Priest recognise the particularly ontological flavour of this contradiction:

Nāgārjuna’s enterprise is one of fundamental ontology, and the conclusion he comes to is that fundamental ontology is impossible. But that is fundamentally ontological conclusion — and that is the paradox. (p. 15)

This is the ontological paradox and Garfield and Priest attribute it to Nāgārjuna. According to Garfield and Priest, because ultimate reality is contradictory in this way, Nāgārjuna can assert ultimate truths corresponding to it, even though there are no ultimate truths for him.

Is Emptiness Contradictory?

The crux of their argumentation is the inference from the lack of essential nature (i.e., emptiness) to emptiness being the nature of all things, as they argue: ‘To be ultimately empty is, ultimately, to lack emptiness. In other words, emptiness is the nature of all things’ (ibid.). Is there any reason to think that Nāgārjuna would accept this inference? It is hard to imagine that he would. To articulate the reason why that must be the case, we have to delve into the semantic principles that Nāgārjuna seems to presuppose (at least if we understand Nāgārjuna in the way that Garfield and Priest do). Garfield and Priest argue that there are ultimate and conventional truths corresponding to ultimate and conventional realities. Moreover, emptiness is empty because ‘by virtue of [[[emptiness]]] [all things] have no nature, not even emptiness’ (ibid.). So emptiness functions as a truth-maker for the claim that everything, including emptiness, is empty. It is emptiness

that makes it true that, or in virtue of which, emptiness is empty. This suggests that Garfield and Priest are assuming the truth-making principle: there must be something that makes a truth true. That is to say that a statement is true if there is something that makes it true. Priest explicitly accepts this principle in the context of embracing Nāgārjuna’s paradoxes:

I am enough of a realist to hold that there must be something about reality that makes [statements true]. ... When I say that reality is contradictory, I mean that it is such as to render those contradictory statements true. (Priest (2002): 295)

The orthodox semantic account in Indian philosophical circle might have been the truth-making principle as Garfield (1996) and Westerhoff (2009, ch. 9) suggest.xii In this respect, Garfield and Priest might be following this historical account. However, an attribution of the truth-making semantic principle to Nāgārjuna (whether or not such an attribution is legitimate) doesn’t allow them to present a coherent reason why Nāgārjuna would assert ultimate truths after denying their very existence. In subscribing to the truth-making semantic principle in the context of embracing Nāgārjuna’s paradoxes, Priest explicitly accepts realism that accompanies the truth-making semantic principle as we can see in the passage quoted above. That the truth-making principle presupposes realism has been wellrecognised. As Armstrong (2004) puts it:

To demand truth makers for particular truths is to accept a realist theory for these truths. There is something that exists in reality, independent of the proposition in question, which makes the truth true. ... It is in virtue of that independent reality that the proposition is true. What makes the proposition a truth is how it stands to this reality. (p. 5) 

So if we assume that Nāgārjuna (implicitly) adopted the truth-making principle to establish the truth of emptiness, he would have to been seen as endorsing the realism about emptiness. But emptiness is the absence of svabhāva. We would then have to think that Nāgārjuna was arguing that emptiness is empty in virtue of the absence of svabhāva. So, if we were to think that emptiness functions as the truth-maker for the ultimate truth that emptiness is empty, Nāgārjuna would have to be seen as a realist about absence. The Naiyāika might be a realist about absence (abhāva) and can accept the ‘positive’ quality that absences have.xiii But how can we think of Nāgārjuna as being a realist about absence? If they wish to show that Nāgārjuna would embrace the ontological paradox (and the paradox of expressibility), Garfield and Priest would have to establish that Nāgārjuna was a realist about absence. Without doing this, they can’t show that Nāgārjuna’s doctrine of emptiness would entail the ontological paradox.xiv Another way to show that Garfield and Priest’s attribution of ontological paradox to Nāgārjuna is dubious is to examine their response to the suggestion that Nāgārjuna asserts only conventional truths. Given that the assertion of ultimate truth contradicts the claim that there are no ultimate truths, the suggestion would save Nāgārjuna from contradiction. In response, Garfield and Priest argue:

When, for instance, a mādhyamika says that things are ultimately empty, that claim can be cashed out by saying that when we analyse that thing, looking for its essence, we literally come up empty. The analysis never terminates with anything that can stand as an essence. But another way of saying this is to say that the result of this ultimate analysis is the discovery that all things are empty, and that they can be no other way. This, hence, is an ultimate truth about them. (p. 12)

Similar to the way they infer an existence from an absence, Garfield and Priest infer the existence of an analysis that never terminates from the absence of an analysis that terminates. In order to recognise that

an analysis that never terminates exists, we would have to have a higher-order perspective from which we can recognise that an analysis never terminates. Consider, for instance, the successor function for natural numbers (s(n) = n + 1). This function would generate a sequence of natural numbers. To recongnise that the sequence would never terminate is to think that there exists a theory (of natural numbers) which is closed under the successor function. What make it a theory, however, are not only the formula itself but also the terminating clause which says “that’s all and nothing else” allowing us to make claims about the totality of natural numbers. It is this clause that makes a set of formulas a theory (which is closed under some operation) rather than just a formula. But the existence of a formula doesn’t entail the terminating clause. Indeed, it is exactly such a theory construction as a result of adopting a higher-order (or ultimate) perspective that Nāgārjuna rejects by means of his doctrine of emptiness.xv If we were to think that Nāgārjuna was also recommending the theorisation of our analyses by inferring the terminating clause from the existence of some formula and using the resulting theory as an analysis while rejecting exactly that, it would make nonsense of what he was trying to accomplish in terms of emptiness. We would have to think that whatever he rejects, he also accepts it. This would make Nāgārjuna not a dialetheist (someone who holds that some contradictions are true) but a trivialist (someone who holds that all contradictions are true). This would reduce Nāgārjuna to our imaginary Whitehead and Russell. Moreover, it is difficult to see Nāgārjuna as accepting the existence of a totality of which all empty things are parts. In describing Nāgārjuna’s paradoxes, Garfield and Priest introduce the Inclosure Schema whose first clause is

Ω = {x : ϕ(x)} exists, and ψ(Ω).

In the context of Nāgārjuna, it says that there is a totality of things that are empty and everything in this totality has a common nature (i.e., emptiness). By claiming that the totality is and isn’t part of the totality, Garfield and Priest derive the ontological paradox. But the first half of this clause presupposes the Domain Principle according to which ‘whenever there are things of a certain kind, there are all of those things’ (Priest (2002): 280). By interpreting the relationship between the empty things and the totality of them as a set membership, Garfield and Priest interpret the claim that everything is empty (of svabhāva) to mean that, for everything that is empty, there is a totality of which it falls under. It is not clear how Nāgārjuna would understand the set membership relation. Nevertheless, if we understand it in terms of a mereological relation (part-whole relation), the Domain Principle entails that there is a whole of which every empty thing is a part. So, if we were to think that Nāgārjuna would accept the Domain Principle, we would have to think that he would affirm the existence of the whole over and above the parts. Again, I fail to see how one could show that Nāgārjuna would accept the existence of the whole over and above the parts. It is important to notice here that Garfield and Priest have to provide an argument why Nāgārjuna would assert ultimate truths. Given that Garfield and Priest argue that there are no ultimate truths for Nāgārjuna, they can’t simply assume that he also assert ultimate truths without trivialising Nāgārjuna’s position. If they want to claim that Nāgārjuna also expresses ultimate truths, they need to present a reason why he would do so. The reason they seem to provide is that ultimate reality is both empty and not empty. As we just saw above, Garfield and Priest present an argument that crucially relies on some inferences and principles. But there is no reason to think that Nāgārjuna would accept them based on other things that Garfield and Priest claim he would accept. Hence, Garfield and Priest haven’t established, without begging question, that Nāgārjuna expresses ultimate truths that can’t be expressed. Thus, they haven’t shown that the ontological paradox and, thus, the paradox of expressibility, can be

attributed to Nāgārjuna. They own us a reason ⎯ whether in terms of an ultimate analysis or a conventional analysis ⎯ why we should think of Nāgārjuna as accepting such paradoxes.

Truth-Makers

Now, even though Garfield and Priest have failed to establish that Nāgārjuna would accept the paradox of expressibility and the ontological paradox, there would be a contradiction in the vicinity if we followed their lead. There is a question whether Nāgārjuna would accept the truth-making principlexvi and so I wouldn’t endorse the resulting contradiction on behave of Nāgārjuna. If we nevertheless assumed the truth-making semantic principle and realism that accompanies it, however, we could argue that the doctrine of emptiness entails a contradiction. It is just that a contradiction that would arise is different from the contradiction Garfield and Priest identify. First, consider the Barber paradox. The Barber paradox is about a barber who shaves all and only those who do not shave themselves. As Sainsbury (1995) puts it:

In a certain remote Sicilian village, approached by a long ascent up a precipitous mountain road, the barber shaves all and only those villagers who do not shave themselves. (p. 2)

If we took this statement about the Barber at face value, we would be puzzled about who shaves the barber. Assuming that the barber is a male, the barber cannot shave himself, since he shaves only those who do not shave themselves. But he cannot not shave himself either, since he shaves all of those who do not shave themselves. As Sainsbury suggests, the exotic remote Sicilian village disguises the shallowness of this paradox. This is because there is an easy solution just around the corner: reject the supposition that there is such a barber. In order to offer this as a solution, however, we would have to be able to say that it is ‘really’ the case that there is no such barber. That is what it is to offer the rejection of the supposition as a solution to the paradox. If we were to simply read off the surface, we would be forever caught in an intellectual mess by forever thinking that there is a barber with contradictory properties.xvii By the same token, if we think of Nāgārjuna as accepting the realism of the truth-making principle, we can think of him as saying that it is ‘really’ the case that there is no ultimate truth when he asserts that there is no ultimate truth as a true statement. One may further take it that it is ultimately true that there is no ultimate truth, as Garfield and Priest do. But it is exactly such an assertion that is contradictory. If there is no ultimate truth, then it can’t be an ultimate truth that there is no ultimate truth. And there being no ultimate truth contradicts there being an ultimate truth. Thus, so one might argue, the truth-making principle and the realism that accompanies it entails that Nāgārjuna would have to be committed to a contradiction. But how would such a contradiction really arise? The contradiction is that there is no ultimate truth (that is entailed by Nāgārjuna’s doctrine of emptiness) and also that there is an ultimate truth (because there must be something that makes it ultimately true that there is no ultimate truth). If we think of Nāgārjuna as arguing that there is no ultimate reality based on the rejection of the existence of svabhāva that ultimately gives an object its numerical identity (as Garfield and Priest argue, as we saw above), then we must think of him as rejecting the existence of truth-makers for any ultimate truth. This means that Nāgārjuna is rejecting the applicability of truth-making principle to ultimate truth. Nonetheless, if we think that the semantic principle underlying Nāgārjuna’s doctrine is the truth-making principle following Garfield and Priest, we also have to think that, when he says that it is ‘really’ the case that there is no ultimate truth (assuming that he does this following Garfield and Priest), he is

asserting the existence of a truth-maker. This means that Nāgārjuna is also applying the truth-making principle to ultimate truth. So the contradiction is that the truth-making semantic principle is both applicable and not applicable to ultimate truths. This is not to say that emptiness is both empty and not empty and in virtue of which there are and aren’t ultimate truths. Instead, it is to say that there are and aren’t truth-makers for ultimate truths. If Nāgārjuna’s argument (understood in this way) is successful, there are no truth-makers for the ultimate truths (since there are no ultimate truths). If he wants to say that that is ‘really’ the case by being a realist, then he has to reject it by also claiming that there are truth-makers for ultimate truths. If ‘Nāgārjuna’s enterprise is one of fundamental ontology’ (ibid.) as Garfield and Priest claim, then that may well be how we would have to think. But this doesn’t say that what ultimately exists is contradictory. All it says is that we have to accept the very principle that we reject. If we want to present a view of what ‘really’ exist, we have to investigate it as if there is a truth-maker that makes a certain statement ‘really’ true. But when the investigation demonstrates that there is no such truth-maker, then it is the assumption which we adopt in order to engage with ontology that must be rejected. This may show that the investigation into fundamental ontology involves contradictions. It doesn’t follow, however, that how things exist is contradictory. Of course, if we want to say that it is ‘really’ the case that an investigation into what fundamentally exist involves contradictions, then, following the truth-making principle, we have to accept that there is something that makes it ‘really’ the case. All this shows, however, is that what we do in the name of ontology (‘ontology’ understood as a study of what exist) that is contradictory. From the contradictory nature of our activity, we can’t infer the contradictory nature of the object of investigation. The fact that our investigation necessarily ends up in a contradiction doesn’t show that the object of investigation (what exist) is also contradictory.xviii One way to make this point is to say that the intelligibility of what exist consists of contradictions. Given that we ‘see’ things, we might ask how and why they exist. Any attempt to answer the how and why questions may require us to suppose that the answers can be given in virtue of the things we suppose to exist. That is, if we want to make them intelligible to us, we have to be able to apply the truth-making principle so that we can make sense of how and why they exist. If, in the process, we discover that there is no truth-maker that can be applied, then we have to reject the application of the principle even though this rejection is possible because of the application of the principle. A contradiction might arise in the process of making emptiness intelligible to us given that we have to suppose the applicability of the truth-making principle even thought it is this supposition that would have to be rejected. The contradiction arises, however, not in the object of investigation but in our activity of investigating it. Thus, even if we agreed that Nāgārjuna would have to accept contradictions, he couldn’t be seen as accepting that ultimate reality is contradictory. Contradictions would arise somewhere else. Garfield and Priest might object to this way of understanding Nāgārjuna. If we think that intelligibility is possible only by conventional means (since our cognition requires conventions such as language), it is our conventional reality/truth that may be contradictory. They seem to argue, however, that conventional reality/truth is consistent for Nāgārjuna.xix But if conventional reality/truth is identical with ultimate reality/truth (Garfield and Priest (2003): 6) and ultimate reality/truth is contradictory as Garfield and Priest claim, then conventional reality/truth must also be contradictory even if it is also consistent. Thus, they can’t insist that conventional reality/truth is consistent and, on that basis, reject the suggestion that it is our intelligibility of ultimate reality rather than ultimate reality itself that is contradictory.

The Deflationary Semantic Account

The contradiction that we saw in the previous section arises because of the truth-making principle that Garfield and Priest assume in the context of Nāgārjuna. What if Nāgārjuna’s semantic principle were different? There isn’t much reflection on the semantics by Nāgārjuna himself. So it is hard to pin down what exactly his semantic principle was. But Mādhyamikas seem to have developed different accounts of truth.xx If we think of Nāgārjuna as basing his doctrine of emptiness on a different semantic account, can we provide an explanation as to why Nāgārjuna would assert ultimate truths after arguing that there are no ultimate truths? If Nāgārjuna was a deflationalist as Priest, Siderits and Tillemans suggest,xxi can we still think of Nāgārjuna as propounding contradictions? A deflationalist holds that all there is to truth is the Tschema:

p〉 is true iff p

where p is a proposition (or a statement) and 〈p〉 is the truth-bearer. A deflationlist denies the referential relation, in terms of truth-making, to guarantee the bi-conditional. Given that the bi-conditional doesn’t require the existence of truth-makers, the T-schema itself gives rise to a weaker notion of truth than the truth-making principle. As Priest (2000) argues, the deflationalist account of truth may not be able to avoid contradictions because the T-schema itself seems to give rise to contradictions. What Garfield and Priest need to explain in our context is, however, why we can think of Nāgārjuna as asserting ultimate truths after arguing that there are no ultimate truths. The deflationalist accounts for truth only in terms of the T-schema. But T-schema can’t be used to explain why something is true other than simply pointing at the T-schema. This means that all a deflationalist can say as an explanation of why something is an ultimate truth is: ‘just because!’ If this is the response for a request of an explanation, it is question begging. If this is the case, Nāgārjuna is no better than the fictional Whitehead and Russell who would simply assert that 1 + 1 ≠ 2 while proving that 1 + 1 = 2. Thus, Garfield and Priest can’t show by appealing to the deflationalist account of truth that Nāgārjuna had a good reason to contradict himself. There may be other accounts of truth that can be used to unpack Nāgārjuna’s doctrine of emptiness such as pragmatist or fictionalist account of truth. Whatever the alternative account of truth Garfield and Priest might propose, they own us an explanation for why Nāgārjuna isn’t trivially contradicting himself in asserting ultimate truths, especially the ultimate truth that there is no ultimate truth, after arguing that there is no ultimate truth. If they want to claim that the attribution of the paradox of expressibility and the ontological paradox to Nāgārjuna is legitimate, Garfield and Priest need to provide a non-question begging explanation.

Conclusion

Nāgārjuna’s doctrine of emptiness is primarily an ontological doctrine. It is that everything exists empty of essence (svabhāva). Garfield and Priest argue that the doctrine entails that the way in which things exist is fundamentally contradictory. For them, Nāgārjuna advocates (or would advocate) the ontological paradox that the way in which things exist is fundamentally contradictory. I have shown that Garfield and Priest have failed to legitimately attribute the ontological paradox to Nāgārjuna. I have done so by examining the semantic principle that they attribute to Nāgārjuna and by showing that the principle doesn’t allow the fundamental way in which things exist to be said to be or asserted as contradictory.

By implication, I have shown that an understanding of the semantic principle that underlies emptiness is crucial. If everything is empty, we must be able to assert it as true. We must then have to investigate the way in which we can do so. Without this, we have no resources to understand what Nāgārjuna is saying. This is not to argue that Nāgārjuna’s doctrine is just a semantic doctrine. It is primarily an ontological doctrine. The two truths (ultimate and conventional truths) that are used to explain emptiness are also primarily ontological. Nevertheless, they require a semantic principle that underlies them and an understanding of this semantic principle is crucial in understanding the doctrine of emptiness that has been expressed by Nāgārjuna.xxii The doctrine of emptiness is that everything is empty of essence. But the scope of the doctrine covers everything that Nāgārjuna says. Thus, if we are to understand the doctrine propounded by Nāgārjuna, we must understand how he could have done so.

References

Armstrong, D.M. (2004) Truths and Truthmakers, Cambridge: Cambridge University Press. Bhattacharya, Kamaleswar (1986) The Dialectic Method of Nāgārjuna (Second Edition), Delhi: Motilal Banarsidass Publishers. Chakrabarti, Arindam, ‘The Unavoidable Void’, Nothingness in Asian Philosophy, JeeLoo Liu and Douglas L. Berger (eds.), London: Routledge, pp. 3-24. The Cowherds (2011) Moonshadows: Conventional Truth in Buddhist Philosophy, New York: Oxford University Press. Deguchi, Yasuo, Jay L. Garfield and Graham Priest (2013) ‘How We Think Mādhyamikas Think: A Response to Tom Tillemans’, Philosophy East & West, Vol. 63, pp. 426-435. Ferrero, Giuseppe (2013) ‘A Criticism of M. Siderits and J.L. Garfield’s ‘Semantic Interpretation’ of Nāgārjuna’s Theory of Two Truths’, Journal of Indian Philosophy, Vol. 41, pp. 195-219. Garfield, Jay L. (trans. and comment.) (1995) The Fundamental Wisdom of the Middle Way: Nāgārjuna’s Mūlamadhyamakakārikā, New York: Oxford University Press. — (1996) ‘Emptiness and Positionlessness: Do the Mādhyamika Relinquish All Views?’, Journal of Indian Philosophy and Religion, Vol. 1, pp. 1-34. — (2014) ‘Madhamaka, Nihilism, and the Emptiness of Emptiness’, Nothingness in Asian Philosophy, JeeLoo Liu and Douglas L. Berger (eds.), London: Routledge, pp. 44-54. — and Graham Priest (2003) ‘Nāgārjuna and the Limits of Thought’, Philosophy East & West, Vol. 53, pp. 1-21. — and Mark Siderits (2013) ‘Defending the Semantic Interpretation: A Reply to Ferraro’, Journal of Indian Philosophy, Vol. 41, pp. 655-664. Kalupahana, David J. (1986) Mūlamadhyamakakārikā: The Philosophy of the Middle Way, Albany: State University of New York Press. Lindtner, Christian (1986) Master of Wisdom: Writings of the Buddhist Master Nāgārjuna, Berkeley: Dharma Publication. Priest, Graham (2000) ‘Truth and Contradiction’, Philosophical Quarterly, Vol. 50, pp. 305-319. — (2002) Beyond the Limits of Thought (Second Edition), Oxford: Oxford University Press. — (2014) ‘Speaking of the Ineffable’, JeeLoo Liu and Douglas Berger (eds.), Nothingness in Asian Philosophy, London: Routledge, pp. 91-103. — and Koji Tanaka, Zach Weber (2013) ‘Paraconsistent Logic’ Stanford Encyclopedia of Philosophy, Stanford: Stanford University Press. Sainsbury, R.M. (1995) Paradoxes (Second Edition), Cambridge: Cambridge University Press.

Siderits, Mark and Shōryū Katsura (2013) Nāgārjuna’s Middle Way: Mūlamadhyamakakārikā, Sommerville: Wisdom Publications. Tanaka, Koji (2014) ‘In Search of the Semantics of Emptiness’, JeeLoo Liu and Douglas Berger (eds.), Nothingness in Asian Philosophy, London: Routledge, pp. 55-63. Tillemans, Tom J.F. (2009) ‘How Do Mādhyamikas Think?’, Pointing at the Moon, Mario D’Amato, Jay Garfield and Tom Tillemans (eds.), New York: Oxford University Press. — (2013) ‘ “How Do Mādhyamikas Think?” Revisited’, Philosophy East & West, Vol. 63, pp. 417-425. Westerhoff, Jan (2009) Nāgārjuna’s Madhyamaka: A Philosophical Introduction, Oxford: Oxford University Press.

  
  
  I would like to thank Jay Garfield, Graham Priest, Mark Siderits and Tom Tillemans for their helpful comments on earlier drafts of the paper.
   i
  English translations of the text can be found in Garfield (1995) (from the Tibetan translation), Kalupahana (1986), and Siderits and Katsura (2013).
   ii
  Given that ‘contradictions’ are often thought to be applicable only to propositions (or sentences, utterances), it may be thought to be a category mistake to describe reality as contradictory. As it becomes clear later, Garfield and Priest’s claim about Nāgārjuna’s emptiness is not that things are contradictory but that the way in which they exist is contradictory. If the use of the word ‘contradiction’ in this context is bothersome, one may replace all occurrences of ‘contradiction’ by ‘inconsistent’. For example, sets, beliefs and methods that are not necessarily linguistic can legitimately be said to be inconsistent even though they may not be described as contradictory. iii
  There are several mystical interpretations of Nāgārjuna and some of them may not be best described as ‘antirational’. See, for example, Lindtner (1986) and Priest (2014). The version of mysticism Garfield and Priest argue against, however, is the one described in the previous paragraph. Given that the focus of this paper is Garfield and Priest’s interpretation, I refrain from articulating other mystical interpretations. iv
  For an introduction to paraconsistent logic, see Priest, Tanaka and Weber (2013). v
  See Tillemans (2013). A reply can be found in Deguchi, Garfield and Priest (2013).
   vi
  See, for example, Westerhoff (2009) pp. 183ff. How exactly to understand Nāgārjuna’s claim about no thesis is a matter of controversy among both traditional and modern commentators. I won’t elaborate on this controversy in this paper since it focuses on a different set of issues. vii
  See Tillemans (2009).
   viii
  For some arguments against mystical interpretations, see, for example, Garfield and Priest (2003) and Garfield and Siderits (2013).
   ix
  Some of Garfield and Priest’s response to these traditional attempts can be found in Garfield and Priest (2003) as well as Deguchi, Garfield and Priest (2013).
   x
  Just like Garfield and Priest, I won’t labour through these arguments in this paper.
   xi
  As Westerhoff (2009) points out, there are three senses of svabhāva that Nāgārjuna employs: essential property, independence, and withstanding any (logical) analysis. See Westerhoff (2009) § 2.1. For the purpose of this paper, how best to understand ‘svabhāva’ is not of importance. As we will see, the importance lies in the semantics of emptiness rather than ontology and, thus, I shall not articulate the different ontological senses of svabhāva that Nāgārjuna invokes in presenting his doctrine of emptiness.
   xii
  The Cowherds (2011, ch. 8) suggest that it might not be best to think of Nāgārjuna or Mādhymikas as presupposing the truth-making principle. See also Tanaka (2014) for a discussion of the semantics of emptiness in Madhyamaka.
   xiii
  See Chakrabarti (2014) for several Indian theories of absence.
   xiv
  In fact, Garfield and Siderits (2013) demonstrate the difficulties of showing that Nāgārjuna was a realist about absence.
   xv
  For example, Mūlamadhyamakakārikā IV: 5.
   xvi
  See the Cowherds (2011, ch. 8).
   xvii
  A mystic may hold that any exercise of our intellectual faculty leads to an intellectual mess. However, we are considering an interpretation of Nāgārjuna which rejects such a mystical interpretation.
   xviii
  Garfield and Priest might reject this. They claim that ‘we can think (and characterise) reality only subject to language, which is conventional, so the ontology of that reality is all conventional’ (p. 13). They seem to think that if our characterisation of reality is conventional, ontology of that reality must also be conventional. They might then reject the claim that one can’t infer the contradictory nature of what exist from the contradictory nature of our investigation. That seems to me to be a non sequitur and I don’t know how they can justify their claim. Many thanks go to Mark Siderits for pointing this out to me (in personal conversation).
   xix
  For example, Garfield (2014) claims that ‘the consistent conventional truth is identical with the paradoxical ultimate truth’ (p. 48). Also, given that Garfield and Priest (2003) try to show that ultimate reality is contradictory without spreading inconsistency to conventional realm, they seem to hold that conventional reality is consistent for Nāgārjuna.
   xx
  See the Cowherds (2011, ch. 8) and Tanaka (2014).
   xxi
  The Cowherds (2011, ch. 8).
   xxii
  Ferrero (2013) has argued against the ‘semantic interpretation’ of Nāgārjuna by Jay Garfield and Mark Siderits. Ferrero seems to be making Garfield and Siderits as saying that the two truths which are used to explain emptiness has only semantic content and have no ontological/metaphysical content. The point Garfield and Siderits are making seems to be the same (or similar) point I am making here rather than making emptiness as an exclusive matter of semantics, however. Garfield and Siderits’ reply can be found in Garfield and Siderits (2013).