Ranganatha Sitaram, Ph.D., Correspondence: ranganatha.sitaram@stjude.org
ABSTRACT
Nagarjuna was a Buddhist monk and philosopher in the second century after Christ. He was one of the central figures in the Buddhist tradition of Indian philosophy. After his works were translated into English and other Western languages, he became known in Western philosophical circles for his philosophy of the Middle Way (Madhyamaka in Sanskrit) based on the notion of emptiness. His treatise, Mulamadhyamaka (MMK) has influenced Buddhist thought in India, Tibet, China, Japan, and other Southeast Asian countries for centuries. In this work, we will attempt to elucidate the relevance of Madhyamaka’s theory of causation to the modern sciences. We argue that the Madhyamaka notion of causation resonates with the network-based concepts of causality in the contemporary sciences, and the adoption of his notion dependent origination may facilitate the integration of multiple levels of scientific explanation and understanding. In this essay, we will first explain Nagarjuna’s philosophy of Madhyamaka with an emphasis on the core concept of emptiness, the notion of svabhava (true nature of objects of the world), and Nagarjuna’s rejection of svabhava. Based on these foundational concepts, we will look closely at Nagarjuna’s philosophy of causation and the notion of causal fields, and our interpretation of the Madhyamaka theory of causation as an interconnected network of events and background conditions forming the causal field. We will then shift our attention to lay out Western philosophical and scientific theories of causation. We will introduce the recent developments in the concept of causal networks. Finally, we will circle-back to relate Madhyamaka’s theory of causation to the modern notions of causation.
NAGARJUNA’S MADHYAMAKA PHILOSOPHY
The foundational concept of Nagarjuna’s philosophical work is emptiness (sunyata in Sanskrit) (Edelglass and Garfield 2009, Westerhoff 2009, Garfield and Edelglass 2010, Westerhoff 2011, Westerhoff 2018, westerhoff 2020, Westerhoff 2024). The concept of emptiness is considered in relation to objects and their properties, events, and concepts of the world. The notion of svabhava’(Sanskrit) is essential to understanding emptiness. Svabhava can be translated into English in multiple ways, as ‘intrinsic nature’, ‘inherent existence’, ‘own-being’, and ‘essence’. Emptiness denotes the absence of svabhava (Westerhoff 2009). One needs to understand the totality of emptiness’s two main conceptual dimensions (Westerhoff, 2009), namely the ontological and the cognitive. The ontological dimension is related to the terms substance, essence, and nature. In Buddhist literature, the ontological dimension is divided further in three distinct ways: essence svabhava, substance svabhava, and absolute svabhava. Essence svabhava is the specific quality of an object that distinguishes it from other objects. It is also described using the essential property of an object in such a way that the loss of that property leads to the cessation of that very object. Westerhoff (2009) gives the example of “heat of fire” as the essential property, as removing heat from the fire leads to the cessation of the fire. According to this notion, “every instantiation of fire is the instantiation of heat (Westerhoff, 2009). On the other hand, heat is not an essential property of water, as water can be hot or cold. Heat is a specific property of water at a given state of water. Other examples are the fluidity of water, the openness of space, the hardness of a rock, and so forth. “Every instantiation of fire is the instantiation of heat” (Westerhoff, 2009). Essence svabhava is an epistemological concept. In Nagarujuna’s philosophy, the ontological notion of svabhava, termed substance svabhava receives greater significance. This notion of svabhava is not about identifying a property of an object that distinguishes it from other objects. Substance svabhava is defined as svabhava existing fundamentally independent of other entities. According to the MMK, this type of svabhavadoes not result from causes and conditions. Substance svabhava indicates the ontological status of objects that do not depend on anything else but exist independently of their own accord. “Substance svabhava is not artificially created and not dependent on anything else” (Westerhoof, 2009).
Substance svabhava is based on the idea that through analysis, we can eliminate what is not real, eventually retaining the basic substance of the object. To identify the substance svabhava of a car, for example, by analysis, we reduce the car into individual parts and reduce individual parts their parts until we can no longer identify further parts of the matter. During this process, we would have dissolved the individual parts upon analysis. Hence, what is considered initially the essence or true nature of the car breaks down and is lost during analysis. The vehicle is no longer independent of its parts, which depend on their particular matter. In this manner, one can realize that objects of the world to which we ascribe conventionally svabhava are, in reality, dependent on their parts and their interactions.
Dependence relations
What is meant by the dependence of an object on other objects? There are two ways of considering dependence: existential dependence and notional dependence. An example of existential dependence is a sprout depending on its seed for its existence, including soil, water, air, and other necessary conditions. Another example is that of a car that depends on its parts for its existence. Notional dependence is related to the labeling or naming of objects and concepts. The notion of North India depends on the notion of South India. If South India ceased to be part of the Indian nationhood, then North India would not be an appropriate label anymore. However, North India does not existentially depend on South India as even if (hypothetically) the South Indian landmass is separated from North India, the latter may still exist. Hence, North India depends notionally on South India rather than existentially. Existential dependence can be understood under three main categories: 1) causal dependence, 2) mereological dependence, and 3) conceptual dependence. The first is based on causes and conditions for existence, the second on the existence of parts in the whole, and the third on a subjective mind which designates concepts, names, and labels of objects, events, and happenings in the world. The notion of substance svabhava has been found in Western philosophy, e.g., by Descarte’s substance being of other objects and Wittgenstein’s notion of logically simple objects.
The rejection of substance svabhava by property argument
Primary existents are objects that possess substance svabhava, the property that does not dissolve with analysis and remains as true nature. In the olden Buddhist tradition, primary existents were called Mahabhutas(i.e., earth, water, fire, and wind). However, Nagarjuna rejects the notion of primary existents. The primary existents are those that have essential properties with which they are identified (e.g., earth with hardness, fire with heat, water with fluidity, etc). Hence, without these properties, primary existents can no longer be identified as such. However, a question that arises is what existed before the primary existents acquired these properties.
If we assume that they existed earlier without any property, then the question arises when they acquired their property. If they indeed acquired their property, they are then dependent on those properties, in which case, they are no longer independent and hence are not primary existents. We are thus in infinite regress by this analysis. Hence, Nagarjuna concludes that there are no independent primary existents. Nagarjuna concludes that objects in the world are dependently originated.
The rejection of substance svabhava by mereological argument. The Madhyamaka contends that, like the previous argument of the non-existence of primary existents, the notion of parts and wholes (mereological) is flawed. In the ordinary sense, objects of the world are considered either complex (made of many parts) or simple (without parts). However, complex objects cannot be primary existents because they depend on their parts. As primary existents are independent of anything, they cannot be dependent on their parts. Considering simple parts-less objects, they must have their unique essential properties. In such a case, the dependency of the part on its property leads us back to the problems of the property argument considered earlier. Hence, according to Madhyamaka, primary existents are neither complex nor simple.
The rejection of substance svabhava by the argument from change
According to Madhyamaka, the notion of substance svabhava is incompatible with change. Substance svabhava implies that a primary existent is independent, irreducible, and not artificially constructed. Primary existents are fundamental constituents of reality. Hence, by definition, primary existents do not change, as do the objects they constitute. This means that the world would be static, which goes against our observation of the world as being dynamic.
The cognitive dimension. To perceive substance svabhava is a natural human tendency, a cognitive default (Westerhoff 2009). According to Candrakirti, a Buddhist monk and philosopher who followed Nagarjuna, one who understands dependent origination will cease to perceive substance-svabhava (Westerhoff, 2009). The rejection of the substance svabhava leads to a clearer understanding of dependent origination. According to Madhyamaka, a correction to the automatic human disposition to perceive substances with svabhava brings about true perception. In Buddhist philosophy, attaching essential properties to objects leads to suffering. Realizing emptiness is achieved by understanding that our minds superimpose properties onto objects in their attempt to conceptualize the world. This is a default trait of the mind that results in automatically seeing objects as possessing intrinsic nature or essence. Madhyamaka suggests two stages of correcting this view. The first stage is to understand by following the Madhyamaka analyses that svabhava does not exist. The second stage involves truly assimilating this concept through the “realization of emptiness” by specific training. Hence, the ultimate goal of Madhyamaka is not just the logical understanding of the rejection of svabhava but the realization of emptiness through a cognitive change that eliminates the automatic superimposition of svabhava.
The Madhyamaka philosophy of causation
Having established the basic tenets of Madhyamaka, we shall now consider how Madhyamaka applies these tenets to explain causation (Westerhoff 2009). Causation is a fundamental notion we use readily in daily life to perceive and make connections between events and happenings in the world. Cause-and-effect relationships are basic ways in which we make sense of the world. Causation is the automatic and deliberate application of cognitive tools in everyday life. Nagarjuna argues that the commonsensical notion of cause and effect as independent and distinct entities is problematic. To give an example of our conventional notion of cause and effect, a seed is a distinct object from a tree that grows out of the seed. The seed exists only until the tree starts to grow. Once the tree has started its existence, the seed’s existence ends. The other conventional notion of causation is that cause and effect exist independently of cognizing subjects. This is called the objectivist understanding of causation, which states that the connection between causes and effects is independent of observing beings. Nagarjuna argues through a series of analyses that the above notion of causation is flawed and unsatisfactory.
Interdependence of cause and effect. In contradistinction to the general notion of cause and effect being independent, Nagarjuna posits the interrelationship and dependency between cause and effect. In carrying out this argument, Nagarjuna includes a third category of entities: the supporting conditions that enable causes to produce effects. E.g., a spark plug in a motor car can start the engine (effect) upon firing the spark (cause) only in the presence of the fuel (condition). In Madhyamaka terminology, the collection of background conditions is called samagri, meaning complete collection. Westerhoff (2009) uses the term causal field to mean the cause, effect, and the collection of all supporting conditions. According to the Madhyamaka, causes and effects are interdependent. An oak tree (effect) does not exist without an acorn seed (cause). However, this change may not come by without the background conditions, including suitable soil, sunlight, water, space to grow, and so forth. The effect depends on the existence of the cause. On the corollary, the cause may not appear to need the effect for its existence. However, a deeper analysis based on Madhyamaka shows that causes also depend on effects. Although this conclusion may appear counterintuitive at first look, Nagarjuna’s presentation of this argument in the context of causal fields makes it clearer. The causal field, as described earlier, is a complete collection of the cause, the supporting conditions, and the effect. When there is no effect, the cause and the background conditions forming the causal field cease to exist. Hence, Nagarjuna contends that a cause depends also on the effect given the causal field’s content, tying together the cause, the effect, and the conditions. Hence, by the foregone analysis, the cause is not just notionally dependent on the effect (by the label given to it due to its link with the effect) but is also existentially dependent on the effect. In summary, the causal field exists only if the effect is produced, which requires the causes and conditions to be ready. Due to this interdependence of causes and effects, it follows that the claim that causes and effects are not independent and exist with their svabhavas is not true. Cause and effect are strongly dependent on each other and hence devoid of their own svabhava.
Four ways of causal production
After establishing that the causes and effects are interdependent, Nagarjuna considers the four different ways in which causes could be produced:
- the effect exists already in the cause,
- the effect is produced by something else,
- the effect caused by itself and something else, and
- the effect is not produced by anything.
We will now consider the Madhyamaka analysis of each of these possibilities.
Rejection of causation as self-causation. Nagarjuna considered two separate views of self-causation. One is that the same object is both the cause and the effect. The seed that causes the tree is not the same as the tree. Nagarjuna rejects this view. Causes are considered to precede effects in time; hence, it is difficult to see how causes and effects can be the same object. The second view, called the mereological argument, is that the effect is already part of the cause. Nagarjuna rejects this view for two reasons. The first reasons is that the effect need not be produced by the cause because it is already part of the cause. As causes are considered to produce effects, the above argument does not hold well. The second reason is that if the effect was already part of the cause, then it should be evident in the cause, which is not a fact. For example, knowing all the causes and conditions of a combustion engine does not show us the effect until it has happened, even if we can infer the effect by prior knowledge or estimation. In other words, the literal presence of the effect in the cause is not seen without it happening. Hence, self-causation is not an acceptable view of causation, according to Madhyamaka.
Rejection of causation by another object. The most natural and conventionally understood view of causation is that cause and effect are distinct phenomena. The spark is distinct from the explosion. Nagarjuna uses two arguments to reject causation by another object (Westerhoff, 2009). Let us assume that an object xt was caused by another object xt-1. The objects xt and xt-1 cannot be the same object as this implies self-causation, which has already been rejected. The cause of xt-1 should be an independent object, let us say xt-2. If we proceed in this manner, we can continue ad infinitum to identify a new distinct/independent object as the cause of the last effect in the series. Hence, causation by a series of distinct objects leads to infinite regress. Hence, causation by another object is an unsatisfactory explanation.
Nagarjuna’s second argument against causation by another object is based on the independence of causes and effects, such that causation is not possible at all. This makes it difficult to distinguish between objects that are causally related and those that are not. No two objects in this causal chain can be identical objects as all causes and effects by assumption are distinct objects with their own svabhavas. Note that when objects are distinct and independent in Madhyamaka, they are not just different in their specific svabhava, they are also different in their substance-svabhava. However, the above point contradicts the notion that effects are formed by causes, and hence, the two are causally dependent. This distinction between cause and effect necessarily entails that they stand separate from each other and, therefore, cannot be modified by each other. However, the effect is existentially dependent on the cause, and the cause is at least notionally dependent on the effect. Hence, the causal link between cause and effect is a blatant contradiction, and hence Nagarjuna rejects the view of causation by another object.
Rejection of causation by itself and another object. Following from the first two alternatives of causation, causation by the joint action of the object and another object appears to be easy to rejected. However, the obvious conclusion might not suffice in all cases, as in the case of two chemicals that may not cause an explosion on their own, but their combination can. Nagarjuna rejects this joint causation with a sounder argument. Let us consider the premise that in the example of a statue made of marble by a sculptor, the marble can be considered to have the potential for self-causation in the sense that the statue already exists in the material mass of the marble. However, the statue is not produced until a sculptor works on the marble. Yet, the sculptor cannot work on the statue unless the marble is available. Hence, the marble, the statue, and the sculptor all depend on one another. This implies that the cause, effect, and condition are distinct and independent entities and do not require one another, which contradicts the notion that they are causally connected and dependent on each other. Hence, the third alternative is also rejected by Madhyamaka.
Rejection of absence of causation. Nagarjuna rejects the absence of causation based on epistemic considerations. Our everyday experiences from which we gather facts about the world would not be possible if the world were not causal. It would be unclear then how we make sense of the world. This is Nagarjuna’s first argument against the absence of causation. There are certain entities that might be considered to be outside the causal nexus (e.g., mathematics and abstract knowledge), but Nagarjuna does not seem to be including them in his analysis. The second argument is that the absence of causation would have presumably created a world entirely different from ours. In such a world of non-causality, events, and processes may not follow regularity and order. As we do not experience such a world, this alternative appears hard to defend.
Conclusion. By the above four types of analyses, Nagarjuna shows the flawed conception of objects having svabhava as distinct and independent. However, our default cognitive state as cognizers automatically superimposes svabhava on objects. Nagarjuna points out that this way of viewing the world is demonstrably contradictory to our observation that objects interact with one another and are causally related. Causes and effects depend on each other existentially and notionally. Causation happens in the nexus of interacting objects, events, and supporting conditions, which can be jointly considered the causal field.
CAUSATION IN WESTERN PHILOSOPHY AND SCIENCE
Causation has at least three meanings in the West, perhaps due to the variety of ways the word has been used in the history of philosophy and the sciences (Bunge, 1979). Causation as a category denoting a causal bond or connection can be shown with the example statement “sparks cause fires. Causation as a principle indicating the general law of causation can be seen in the statement “the same cause produces the same effect.” The causal doctrine as a universal validity of the causal principle is seen in the illustrated “everything happens because of a reason.” A fundamental question that has occupied Western philosophers is whether causation denotes an epistemic or ontological process. Is the causal connection real, or is it just a state of our minds? Is it just what is generated by the mind, or does it have its existence in the reality of the universe?
The philosophical problem of causation considers several clusters of issues that Bunge (1979) classifies under two major subsets. The first is the ontological problem, which asks questions such as: how can causal links be characterized, are causal links part of the real world, are there laws governing causation, what place does chance have in causation, and how are they related, and how do they interact with each other? The second is the methodological problem of causality, which is related to the criteria for determining causation, the way to recognize a causal link, and the methods for testing a causal hypothesis.
Aristotle stated the earliest definition of causation, which prevailed during the Renaissance. He described four kinds of cause (Aristotle, Metaphysics), namely, 1) the material cause, standing for the matter or raw material composing the object(s) in question, 2) the formal cause, representing the essence or the quality of the object(s), 3) the efficient cause, the external force that produces the change in the object(s), and 4) the final cause, the ultimate reason or the goal for which the objects and events act. The modern thought on causation started with Galileo’s definition of causation, which eliminated the material, formal, and final causes and retained only the efficient cause as: “that and no other is to be called cause, at the presence of which the effect always follows, and whose removal the effect disappears (Galileo, Il Sagiatore, Bunge, 1979).” Galileo’s definition of cause is not considered correct anymore (Bunge, 1979) as it is too general and prescribes necessary and sufficient conditions for any process, even if it is not causal. The empiricists Locke, Berkeley, and Hume regarded causation as a mental construct identifying the connection between the cause and its effect. They and their followers held that causation connects experiences rather than things in themselves. Hume held that the experienced event called cause produces the experienced event called the effect, but these events are not out there in the world, nor are the connections that our minds infer. Hence, causes and effects are ideas (mental artifacts), are epistemological, and have no ontological bearing. The argument against this notion is that causation is not a mental construct but a factual connection or trait of the external and internal world and has an ontological basis. According to the Genetic Principle (Bunge, 1979), everything depends on specific conditions and happens only when these conditions are met. The Roman philosopher Lucretius (58 BC) held that “nothing comes out of, and nothing passes into nothing. There are neither absolute beginnings nor absolute terminations, but everything is rooted in something else and leaves, in turn, a track in something else.” Does the above mean that things lawfully follow from previous things? The Principle of Lawfulness (Bunge, 1979) states that nothing happens unconditionally and lawlessly. The genetic principle and the principle of lawfulness together form the principle of determinacy. This principle states that everything is determined by something else following specific laws, the something else being the internal and external conditions of the object.
In the history of philosophy, there have been different views on causation: Causalism, Semicausalism, and Acausalism.
Causalism: According to the rationalists, the causal principle is a necessary outcome of thought. It is worked out a priori but not by any scientific analysis. In other words, the causal principle is given to us by the way the mind works. According to Immanuel Kant, the causal connection is “synthetic” because it is verifiable by experience but cannot be derived or refuted by experience (ref).
Semicausalism: This view states that causation is one of many different ways in which events and processes are interrelated and connected. In this sense, causation has scope in the way the world works.
Acausalism: Acausalism is based on the idea that events follow one another in a temporal sequence. These events sequence. These events may have statistical correlations (based on Pearson), or probability laws (Reichenbach), or may follow functional laws (Mach). However, the law of causality is a relic of the past, like an old idea that is no longer relevant (Russel). An extreme doctrine of acausalism is the indeterminist doctrine that rejects all lawful links between events and processes of the world, hence denying causal connection altogether. According to this view, events happen spontaneously. Bunge (1979) suggests that this extreme view has not been defended by later philosophers.
Bunge is against the notion that causation is the only category of determination of events in the world and has unlimited validity. Instead, he favors semicausalism, which posits that causation is one of the different paths to determination, including statistical, teleological, and dialectic determination.
The state-space approach to causality. In the modern philosophical and scientific developments of causation, the state-space approach has been a prominent way to denote and analyze causation systematically (Bunge, 1979). The state-space approach could be applied to objects and events in the world, such as physical, chemical, or biological systems, ecosystems, or social systems. Causation can be described by assuming at each instant that the system is at a specific state relative to a reference frame and does not remain in the same state forever. Each state of the system, x, can be described by the list of all known system properties as a mathematical function. The list of n-properties (called the n-tuple) is a state function of the system, F. The time-dependence of the system is represented by F(t), so that the system moves with time t in an abstract space. This space is called the state-space of the system x, denoted by S(x). A change in the system is denoted by an ordered pair of points in the state-space S(x) and visualized by an arrow joining the two points. The set of all such pairs of states is the event space of x, E(x). The event space E(x) can be obtained by the cartesian product of S(x) by itself. To say that any event e belongs to the event space, we use e 
The modern development of causal models has forged cross-disciplinary progress in estimating causality in mathematical, computational, physical, biological, and medical sciences (Pearl 2000), (Pearl and Mackenzie 2018). Causal models represent mathematically the causal relationships between different entities in systems and populations. Philosophers have used causal models to analyze concepts such as actual causation, decision theory, and counterfactuals. A good model can help make accurate predictions about the behavior of a system and the effect of interventions or intentional changes or perturbations to the system. Logic, probability theory, graphs, and mathematical notations form the foundational bases for mathematical models. A variety of causal models have been developed to answer scientific and philosophical questions. Some prominent causal models are deterministic structural equation models, probabilistic causal models, and causal networks.
THE CORRESPONDENCE BETWEEN MADHYAMAKA AND WESTERN CAUSATION
Nagarjuna’s philosophy of causation may resonate with Galileo’s concept of causation, although the notions of svabhava, emptiness, and dependent origination cannot be found in Galileo’s explication of causation. His view of causation as a universal causal interconnection (Bunge, 1979), where an event is an outcome of several events that jointly bring about the effect, parallels Nagarjuna’s notion of causal field (Samagri), where causes, effects, and supporting conditions interact and work in unison. However, Galileo has been criticized by modern philosophers (e.g., Bunge, 1979) on the grounds that the definition includes an indefinite number of causes and events that could make some difference to the effect. Based on the argument that indeterminateness and haziness are inconsistent with causal determinacy, Galileo’s concept of causation has now been rejected by modern philosophers.
I will argue that the modern notions of Causal Networks (Shafer 1995), (Defilippo, Giorgi et al. 2024) come closest to the spirit of Nagarjuna’s causation. Causal networks are a robust framework for modeling causal elements of a complex system (Defilippo et al., 2024). Causal networks model relationships in a system among the elements of the system while considering the influencing factors and variables. Significant developments in the mathematical formalism of causal networks provide direction to the connections between components of the causal networks. In comparison, Nagarjuna’s notion of the link between causes, effects, and supporting conditions within a causal field parallels the philosophical conception of causal networks. He also analyzed the causal and notional dependencies and time dependencies between causes and effects. However, the directional dependencies of the different elements of a causal network were not of great concern in Madhymaka’s causation.
In conclusion, our analysis of the Nagarjuna’s Madhyamaka philosophy causation and the Western philosophical and scientific developments in causation may have very little common ground. However, we believe that the Madhyamaka philosophy of causation has implication for its notion of substance and essencesvabhava, its arguments for the rejection of svabhava leading to the insight into emptiness and dependent origination of world. As Garfield suggests (Garfield 1994, Garfield, 2001), the understanding and assimilation of the principles of dependent arising in Western philosophy may contribute to the integration and understanding of multiple levels of analysis. Levels of analysis are the bedrock of scientific research and explanation, but because of the specialization of individual scientific disciplines, each level is considered in its isolation during scientific investigation. The levels are seldom integrated in forming the overall scientific explanation. The Madhyamaka philosophy of dependent arising might enable a fluid way to move between levels as much as within each level of analysis.
References
Defilippo, A., F. M. Giorgi, P. Veltri and P. H. Guzzi (2024). “Understanding complex systems through differential causal networks.” Sci Rep 14(1): 27431.
Edelglass, W. and J. L. Garfield (2009). Buddhist philosophy : essential readings. Oxford ; New York, Oxford University Press.
Garfield, J. L. and W. Edelglass (2010). The Oxford handbook of world philosophy. New York, Oxford University Press.
Pearl, J. (2000). Causality: Models, Reasoning and Inference. Cambridge, Cambridge University Press.
Pearl, J. and D. Mackenzie (2018). The Book of Why: The New Science of Cause and Effect. New York, Basic Books.
Shafer, G. (1995). “Causal diagrams for empirical research – Discussion.” Biometrika 82(4): 699-700.
Westerhoff, J. (2009). Nāgārjuna’s Madhyamaka : a philosophical introduction. Oxford ; New York, Oxford University Press.
Westerhoff, J. (2011). Reality : a very short introduction. Oxford, Oxford University Press.
Westerhoff, J. (2018). The golden age of Indian Buddhist philosophy. Oxford, Oxford University Press.
westerhoff, J. (2020). The non-existence of the real world. new york, Oxford university proess.
Westerhoff, J. (2024). Candrakīrti’s Introduction to the Middle Way : a guide. New York, NY, Oxford University Press.
Eclectic Musings by Ranga Sitaram 