Free will and Buddhsim

Responsibility is an interesting phenomenon. Whilst, on the surface, taking responsibility merely means declaring that something is within oneโ€™s own control1, its consequences run deep. Often seen as an act of virtue in Western culture, taking responsibility is connected to the idea of what a fulfilling life should look like. Following the American dream, for instance, anyone can achieve anything if they are only willing to put in the necessary work. If you are poor, you wanted it that way. If you are rich, you worked hard. In both cases, it is your responsibility of which life you want to live. The American dream is compelling. Not only is it motivating, it also promises hope for the poor, and legitimizes the rich. Traces of the American dream can be found in many parts of Western culture. Movies such as โ€œThe pursuit of Happinessโ€2 portray it as a rewarding journey that anyone can undertake, only if they work hard enough.

However, like every dream, the American dream is only that โ€“ a dream. In this case, one with little grounding in reality2. At best, itโ€™s an ideal to strive towards, and at worst a gaslight to get confused by. It is this confusion with reality, that turns this potentially useful fiction into a harmful delusion. Arguments about topics like the American dream often devolve into ones about veracity, with a binary outcome: True, or false. As humans, we tend to like these black and white outcomes. They are easy to remember and useful for quick reasoning. However, this encoding of reality is lossy. Information such as context goes missing in the pursuit of clean cut answers. For instance, the American Dream might very well be factually false, but living as if it were true may still be beneficial, making it conventionally true.

The usefulness of this distinction between factual and conventional truths, has been recognized by early Buddhism as the two truths doctrine. In his book โ€œBuddhism as a philosophyโ€, Mark Siderits3 explains the difference between them like so: โ€œconventionalโ€ truths are limited to everyday life, while โ€œultimateโ€ truths correspond to solely facts [p. 74]. In the example of the American Dream, it may be ultimately false that working hard is the primary reason for success, but conventionally it is still true because practicing it generally leads to more success and is thus beneficial. The two truth doctrine is an epistemic theory of truth, which combines โ€œday to dayโ€ truths with facts about the universe. A statement like โ€œif you work hard, you will succeedโ€, is factually not true, but acting as if it were, would still beneficial, therefore making it conventionally true.

This two-valued conception of reality can be applied to other philosophical debates as well. For example, biblical stories, which are often perceived as fairy tales and mere fiction by modern society, can contain deep philosophical insights. When analyzing them for their factual value, it becomes apparent why people perceive them as untruths: Advancements in science and empiric observations prove, that the world was not created in seven days, and neither were women created out of a manโ€™s rib. The Buddhist approach, via the two truths doctrine, could preserve their wisdom. When conceiving stories not as facts but as a medium to encode wisdom, thatโ€™s understood by anyone and that can be transmitted without physical matter, they would indeed be ultimately false, but also conventionally true. A finer quantization of truth, could prevent throwing out the baby (wisdom) with the bathwater (medium of transmission). Would it uncover countless wisdoms, which were previously stepped over in the pursuit of factual truth? I doubt it. Religious stories, while indeed containing conventional truths, also have their dependencies. These dependencies may be other stories, moral values, or doctrines, without which these stories simply do not work. For example, the creation of the world in seven days would not work without a belief in god, rendering the story incomplete. Furthermore, apart from fundamentalists, religious stories are mostly already understood to be true metaphorically, and not factually. But still, the explicit distinction between factual and conventional truths, could prevent a lot of confusion that happens in debates on (religious) stories. Expressions such as โ€œI donโ€™t believe the story of xyโ€ could be split up into โ€œI donโ€™t believe in the fact that xy really happenedโ€ and โ€œI donโ€™t believe in the conventional truth that xy encodesโ€, making it clear whatโ€™s being criticized.

Another possible application of the two truths doctrine, lies in the debate on free will, which is an integral part of Western philosophy. This debate can be broadly divided into three standpoints: Determinism, libertarianism and compatibilism. Determinism argues that free will does not exist because everything is pre-determined. Libertarianism, on the other hand, argues the that free will does exist and nothing is pre-determined. Finally, compatibilism argues that everything is pre-determined, but free will also exists4. Mark Siderits noted that this debate, although popular in western philosophy, did not take place in Buddhism [p. 110]. However, he suggests his own, compatibilist, argument, which takes advantages of the two truths doctrine.


First off, to make sense of the argument, it is important to understand that the Buddhist conception of what is ultimately real, vastly differs from that, what is seen as โ€œrealโ€ from a Western perspective. In Western conception, most things we interact with are seen as being โ€œrealโ€. For example, a person is โ€œrealโ€ because it exists in space, it can be interacted with, and empiric observations have proven that the concept of a person stays constant over a multitude of observations. The same goes for objects such as cars, tooth brushes, concepts such as family, jobs, etc. In Buddhism, however, with the distinction between two types of truth, the idea of something being โ€œrealโ€ has to be differentiated. One might expect that most things we see as real in the West are also seen as ultimately real by Buddhists, but that is not the case. Actually, hardly any things are. For instance, concepts such as feelings (Vedana), perception (Samjna), urges (Samskara) or consciousness (Vijnana), which are titled Nฤma skandhas by Buddhists and essentially form the human mind, as well as the human body (Rลซpa) itself, are all seen conventional truths5. Therefore, Buddhists claim, that people, which essentially consists of a mind and body (= the five Skandhas), are also only conventionally real, since ultimate truths are independent of conventional ones. However, only being conventionally true, does not mean, that something is โ€œnot realโ€, it is simply another type of โ€œrealโ€. With this preface out of the way, letโ€™s explore Siderits compatibilist argument on free will.

The argument

Buddhists believe that at the ultimate level, people are understood to be a series of sets of Skandhas, which for every moment, gets extended by another set of skandha, that depends on the previous set [p. 96]. Thus, the series of sets of Skandhas is deterministic. To fit free will into this picture, Siderits utilizes the doctrine of two truths. Starting from the Buddhist assumption, that people are merely conventionally true, he concludes, that the concept of responsibility is also just a conventional truth, since judgements of responsibility can only be made about people. And with responsibility now being merely a useful fiction, he concludes that free will, which depends on the concept of responsibility, is also just a conventional truth [p.111 - 112]. However, viewing free will as a conventional truth, does not mean that it does not exist or is incompatible with determinism. Instead, free will, can be understood as a useful fiction, which facilitates cohabitance between people, since it implies that people bear responsibility for their actions, and can be held accountable for them. As such, it provides an extrinsic motivation to not act egoistically, and lays the groundwork for a moral system.


Necessary abstraction layer

What Siderits does not talk about, however, is the applicability is this conception of morality in the real world. After all, for every situation one is confronted with, one should simultaneously act in a way that conforms with the conventional truth of morality, while keeping in mind that it is not an ultimate truth, to avoid existential suffering. Would it even be possible to live like that? According to Buddhism, enlightened people (arhats) can. This is because they are released from the cycle of rebirth (saแนƒsฤra), and can see the world as it is, without being confused by conventional and ultimate truths. However, becoming enlightened, is a long path, that can easily span over multiple lifetimes. So, is the clear understanding of free will reserved for only the enlightened ones? While, from the Buddhist perspective, true understanding of ultimate and conventional truth is a trait, only Arhats have, I would argue that normal people certainly can understand it as well, albeit not as consistently. To exemplify what that could look like, it is worth taking a look at the book โ€œMoral Tribes: Emotions, Reason and the Gap Between Us and Themโ€ by Joshua Greene 6. His theory is, that morality is not something inherent to humans because they are โ€œspecialโ€, but simply because acting in a moral, cooperative, manner, increases chances of survival [p.24]. As a result, through the process of Darwinian Evolution, only cooperative humans survived, creating the illusion of shared values being something inherent to their nature [p.189]. Although Greene never mentioned Buddhism, or the idea of two truths, his recognition of morality as just a tool for survival, and not a trait thatโ€™s inseparable from human nature, reflects exactly what Buddhists would describe as conventional and ultimate truth. Furthermore, Greene suggested a way to apply this knowledge to everyday life [chapter 5], which might also be useful for Siderits argument, by utilizing the dual process theory suggested by Daniel Kahneman in his book โ€œThinking, Fast and Slowโ€. According to Kahnemann, there are two different ways of thinking, which, as the title of the book suggests, are fast on the one hand, and slow on the other hand. Furthermore, these opposite ways of thinking can be modelled as two different decision-making systems in the human brain: Emotional or rather automatic thinking is represented by System 1, and manual thinking, which is also more logical, is being represented by System 27. By utilizing these different modes of thinking during every day live, Greene suggests one could differentiate effectively between morality as a fiction and an actual habit, one can incorporate into ones own life [p. 133]. Whether this approach would work for an entire population, is doubtful. Not everyone has the time, mental capacity, or interest to dissect reality into these two layers. Therefore, it creates the risk of a two-level society: Those who are aware of ultimate truths, and those who only know conventional ones. And since knowing conventional truths, like for example moral rules, is generally enough to live a normal life, few will feel the need to learn anything beyond that.


Although this conception of morality as a useful fiction helps avoid confusion, it is unclear as to how useful this fiction actually is. According to Greeneโ€™s theory, its usefulness is proven by the Darwinian advantage it provides. In Buddhism, however, the usefulness of a useful fiction, or conventional truth, is hard to prove. Since most things are understood to be conventional truths, and only very few ultimate truths exist, how would one separate a good conventional truth from a bad one? Not all conventional truths are equally useful, some might even be harmful. To see how conventional truths might be harmful, letโ€™s explore the example, Siderits gave in his book to explain his compatibilist argument: Someone is sleeping in their bed, and another person sneaks into their bedroom to steal all the money out of their wallet, and spends it partying all night. With Siderits argument, the thief would be conventionally guilty for stealing the money, since they deliberately chose to do so. At an ultimate level, however, the thief never had a choice but to steal the money, since, as a person, they are just a deterministic series of Skandhas [p. 111].

Now, letโ€™s expand on this example. Suppose the victim finds out who stole their money the following day, and confronts that person. The confrontation escalates, resulting in the thief killing the victim out of fear of going to prison because they have a criminal record. In the end, the thief, now murderer, gets caught and sentenced to death for their actions. In this example, again, the thief would only be conventionally responsible for their actions. Whatโ€™s different, however, is the consequence. While in Siderits original example, no consequence other than taking responsibility was described, in this example, this responsibility would result in death. In this case, death as punishment for an action, that the thief ultimately didnโ€™t have any choice but to do. If one were to turn back time to the point where the thief walked into the victimโ€™s room to steal their money, everything would happen in the same way again. On a conventional level, the execution punishes the murderer for his actions. Ultimately, though, the execution is not being carried out to punish the actions of the thief himself. Instead, it is an act of consistency, to uphold the moral system the society believes in and benefits from. After all, why would anyone attend to rules, if no punishment would be given for breaking them?

Is this moral system moral?

Therefore, arises the question: Is it moral to punish someone for their actions, which they ultimately had no free will in because, on a conventional level, it is beneficial? From a utilitarian perspective, it may seem so. Borrowing Siderits conception of morality as a โ€œset of psychological adaptations that allow otherwise selfish individuals to reap the benefits of cooperationโ€ [p. 185]3, which has proven itself to be useful through the process of Darwinian Evolution, one could argue that morality has indeed proven its usefulness. However, these arguments assumes that the collective interests should be placed over individual ones. That is because to increase the groupโ€™s chances of survival, few, ultimately innocent, people would have to bear the consequences in the form of legal punishment. Furthermore, some will even lose their lives, like just exemplified, to uphold this moral system. Therefore, Siderits argument presupposes that group interests should be favored over individual ones.

Letโ€™s step back a bit and define what a โ€œcriminalโ€ actually is. Ultimately, a criminal is just a series of Skandhas, like any other โ€œpersonโ€. The title โ€œcriminalโ€ is merely a designator and not an ultimate truth. In this case, it is a designator for the characteristics, which, at a conventional level, are agreed upon to be โ€œcriminalโ€. Examples for this might include stealing, murdering, etc. Therefore, the only differentiating factors between a normal person and a criminal, are the conventional truths attached to them. These conventional truths, like for example โ€œcriminals are dangerousโ€, are not ultimate ones. Instead, they stem from sayings, experience, stories, and other artifacts of society. To highlight the relativity of conventional truths, one can imagine a planet like the earth, which is inhabited by a brutal version of humans. Suppose cooperation decreased chances of survival of the species, and conflict increased it. On this planet, albeit hard to imagine, an entirely different moral ruleset would emerge. For example, since killing someone is beneficial for the species, it would be seen as a morally acceptable act. Ignoring the high probability, that a society as we know it probably would not even be possible on such a planet, their social moral system would be, just as their ultimate truths, inverse to ours. Behavior we see as normal, like not going after someone with a weapon at first sight, would be immoral because it is harmful for the species. The above example shines a different light on the convenient designator โ€œcriminalโ€. On a conventional level, criminals seem like an evil, which has to be brought under control through law and punishment. On the ultimate level, however, their existence bears, like the existence of any โ€œpersonโ€, no intent at all. Any meaning, fate, or role is a product of human thinking thatโ€™s being projected onto them, and simply a convenient designator. Or in the words of Max Horkheimer: Everything, that we perceive, โ€œbears the stamp of editingโ€8 [p. 175] onto itself, implying that it is man made.

Therefore, the treatment of criminals, which would follow from Siderits compatibilist conception of free will, is paradoxical. On one hand, they would be seen as an unwanted part of society, which should be minimized. On the other hand, however, they would be implicitly needed by this very society, to uphold the moral system of values. The only way for this system to work, morally, would be a world, in which everyone agreed to this contract. However, that required everyone to consistently understand reality at the ultimate and conventional level, i.e., a society of Arhats. On one hand, those arhats, who took on the role of the criminals, would be aware of their necessity for society, as well as their unpleasant path. On the other hand, normal arhats would still treat them as criminals at a conventional level, while being conscious of the ultimate fact, that the concept of a criminal is merely a useful fiction. So, this society of Arhats, would be able to live accordingly to the compatibilist argument suggested by Siderits. However, a society of Arhats is a great requirement. After all, would a society of arhats even need conventional truths, as they are aware of ultimate truths, and do not rely on the conventional ones?


The debate on free will is an integral part of Wester philosophy. It is deeply connected with applied ethics, as different positions open or close the door for responsibility, and therefore morality. In Buddhism, however, this debate has hardly been held. Therefore, Siderits compatibilist argument for free will promises to contain novel ideas. Siderits argument builds on top of the Buddhist doctrine of two truths, which proposes two mutually exclusive categories of truth: conventional truths, which serve as useful fictions, and ultimate truths, which are fundamental facts about reality. Siderits argues, that free will is a conventional truth, thus not ultimately real. By understanding free will as a useful fiction, he argues, it can still be used as a basis for morality, since it does not conflict with the ultimate truth that is the deterministic nature of the universe. The analysis of Siderits argument yielded multiple interesting insights. First off, the Buddhist doctrine of two truths, seems to be a universal concept, which has also been recognized by Western philosophers, independent of Buddhism. Examples of such are Joshua Greene, who in his book โ€œMoral Tribesโ€6 described morality as a useful set of adaptations, which is making to the notion of useful fictions described by Siderits. Furthermore, in critical theory, a similar conception of reality can be found. Ideas like โ€œThe constitution of the world takes place behind the backs of individuals and yet is their workโ€9 reflect the Buddhist notion, that only very few ultimate truths exists, with most concepts merely being conventional truths. Furthermore, it highlights, that it is easy to confuse both types of truth.

This risk of confusion is also a point of criticism for Siderits argument, as it requires an extra abstraction layer during every day life. A possible solution for this problem might be to apply Greeneโ€™s idea of consciously utilizing System 1 and 2 thinking, depending on the context of the situation. This, however, also poses a significant overhead compared to just thinking in conventional truths. Another point of criticism has been the difficulty of determining the usefulness of these so-called useful fictions, as it is not clear how to measure it. Finally, the question arose, if the moral system, which would emerge from Siderits compatibilist argument, would even be moral itself. Since, without a common understanding of ultimate and conventional truth, which would equate to a society of arhats, individuals would be at risk of getting treated unfairly for of society.


  1. โ†ฉ

  2. Exploring american dream โ†ฉ โ†ฉ2

  3. โ†ฉ โ†ฉ2

  4. โ†ฉ

  5. Skandha Wikipedia โ†ฉ

  6. โ†ฉ โ†ฉ2

  7. โ†ฉ

  8. Max Horkheimer: Traditionelle und kritische Theorie. In: Ders.: Gesammelte Schriften, Band 4: Schriften 1936โ€“1941. Frankfurt am Main 1988, โ†ฉ

  9. Herbert Marcuse: Philosophie und kritische Theorie. In: Ders.: Kultur und Gesellschaft I. Suhrkamp, Frankfurt am Main 1965, S. 102โ€“127, hier S. 119. โ†ฉ