Free Will and Moral Responsibility in Islamic Philosophy

Document Type: Research Paper


Assistant Professor, Islamic Culture and Sciences Academy, Qom, Iran



According to a common view among Muslim philosophers, a moral agent has free will if and only if she is able to do an action when she wants to and is able to avoid it when she wants otherwise. Implicit in this view is the Principle of Alternative Possibilities (PAP). On the other hand, according to this view, free will is dependent on requirements such as conception, judgement, tendency, decision, and personal volition. In this paper, I discuss the bilateral defects of this view and present a reformed view on free will and moral responsibility in relation to causal predetermination in an Islamic context.



Muslim thinkers have not given separate treatment to the issue of moral responsibility. In sporadic discussions about the origin and extent of human responsibility, they have tried to deal with the moral agent’s free will in relation to divine foreknowledge and causal predetermination and have tried to present a compatibilist and coherent understanding of this topic. Thus, since discussions about moral responsibility are approximately co-extensive with those about human free will (Levy and McKenna 2009, 97), the views of Muslim intellectuals on the moral responsibility can be investigated with respect to their views on human free will with regard to divine knowledge, providence, and, in particular, causal determination.

The relationship between causal determination and free will is one of the most frequently discussed issues by Muslim philosophers, theologians, and scholars of the principles of jurisprudence (uṣūl al-fiqh). This is because, on the one hand, the principle of causality is one of the most important bases of human thinking and a foundation of a large part of philosophical systems and empirical sciences, and, on the other hand, human free will is the basis of religious obligations and teachings and divine punishments and rewards as well as human moral responsibility in relation to others. Of course, there are some currents that are not much concerned with such discussions, such as Ashʿarites, who deny a causal relationship between creatures and free will in human actions. In Ashʿarite thought, the necessity of the unity of divine acts means to consider all actions and reactions, including those in the material world, as divine acts, denying any kind of causal relation between created beings. In other words, there is no logical concomitance between alleged causes and effects, and it is only God’s will that has causal efficacy. For example, fire does not cause burning; it is God who causes burning as a result of fire, and it is the simultaneity or succession of fire and burning that has caused the illusion of a causal relationship between the two.

Besides the denial of causal determination among phenomena in the material world, human free will is also deemed meaningless by Ashʿarites. On one account (Qūshchī n.d., 339; Rāzī 1986, 246-9; Taftāzānī 1409 AH, 4:284), human action has two forms: it necessarily comes into existence upon the agent’s will and its non-existence becomes impossible. This is the very belief in determinism. The second form is that, upon an agent’s will, both the existence and non-existence of the action remain within the domain of the agent’s power, and the agent will be able to do or not to do the action. The latter also has two forms:

(1) Both the alternatives of doing or omitting to do an action are within the domain of the agent’s power, and one alternative (i.e., doing the action) takes place without any cause, which is against the principle of causation and deemed impossible by non-Ashʿari theologians and philosophers.

(2) Doing the action is preferred to omitting to do it, as a result of a preponderator (murajjiḥ), which makes the former preferable to the latter. Now we face the question of whether or not this determination is strong enough to make the non-existence of the action impossible. If the answer is in the affirmative, the problem of determinism will arise again; and if the answer is in the negative, then the question will arise of how doing the action has ruled out omitting to do the action, while the probability of the presence and that of the absence of this determination, which respectively makes doing the action necessary and omitting to do it non-necessary, are the same.

However, as just pointed out, it is generally agreed by Muslim scholars that humans have free will, human agents are responsible, and their free will is based upon PAP though Muslim philosophers believe that the necessary and determinate principles of this free action are absolutely undeniable.

Free Will from an Islamic Perspective

A common pre-philosophical assumption is that responsibility depends on the ability to choose from alternative possibilities. Although compatibilists and incompatibilists have different ideas on many subjects, both groups traditionally have believed that an agent’s ability to have acted otherwise is necessary both for moral responsibility and free action. Of course, incompatibilists go on to emphasize that in no determined world can humans act differently from what they actually perform and that they cannot be the real origins of their actions. Thus, moral responsibility is impossible. Compatibilists, to the contrary, argue that humans enjoy the power of choosing and acting alternatively, and therefore they can reasonably be deemed responsible for their actions and decisions. Therefore, it has been very natural to think that humans deserve blame or respect only when they have the possibility to choose another option. Therefore, the agent would deserve blame if he were able to do something to avoid the blame and he did not; and if they did something preferable while they could do something less admirable, they would deserve praise. However, contrary to this common belief, some philosophers have tried to consider responsibility, and basically moral action, as a reality which is independent of the agent’s ability to act differently. Thus, within the past decades, debates on free will and moral responsibility have dealt with the following question: “Do an agent’s moral responsibility and free will presuppose the ability to have acted differently than one has acted?”

In spite of the fact that Muslim philosophers also emphasize causal predetermination, they consider an agent’s free will as follows: “An agent is such that he or she can do a particular action if he or she wants to, and does not do it if he or she does not want to” (cf. Mīr Dāmād  2002, 94; Ibn Sīnā 1404 AH, 173; Mullā Ṣadrā 1981, 6:307). They regard the possibility of choosing and doing an alternative action as being effective in the generation of free will and consequently of the moral responsibility of the agent. According to this widely held idea, an agent who has done something unacceptable under conditions such as compulsion, hypnotism, brainwashing, and so forth, is not responsible for her action, because she did not have any other choices. Indeed, she has not acted freely and therefore does not have moral responsibility. Thus, there has been a relatively unchallenged principle widely held by Muslim philosophers:

FW: An agent is morally responsible for what he or she did only if he or she did it freely.

The analysis of this principle results in the well-known principle of alternative possibilities, which seems axiomatic and indubitable:

PAP: An agent is morally responsible for what he or she did only if he or she could have done otherwise or decided differently.[1]

It is clear that the ability of an agent to do differently has been emphasized by many philosophers, including Muslim philosophers, and it is dependent on the agent’s awareness, or the possibility of his awareness (and also of the other conditions of moral responsibility). For instance, a driver whose tea cup has been filled, unbeknownst to him, with a drug which disrupts his nervous system, slowing down his responses to environmental events, will not be responsible for his poor reaction in avoiding a crash with a pedestrian. This is because there is no other possibility for him to avoid the accident; he would be able to avoid the accident if he did not drink the tea or did not driver after drinking, but this was not something he was aware of.

However, a driver who sits behind the steering wheel after a harsh argument while he is still angry, knowing that he may have an accident as a result, and yet drives the car and kills a person in an accident will be responsible for his action, because he was aware of the rational possibility of this event and was able to refrain from driving amid his tensions.

Anyway, according to the two aforementioned principles, the moral responsibility of the agent depends on freedom of action, which in turn requires choosing from among possible alternatives. The availability of an alternative action to the moral agent, and thereby maintaining legal and moral responsibilities is so important that some philosophers have denied prior divine will, knowledge, providence, or any other kind of prior causal determination, due to the possibility of undermining the ability to do an alternative. From this viewpoint, if the agent has only one real option, he is like a train—as Feinberg says in his “open option” theory—that is blocked, on the one hand, from all other paths (Goldman and Kim 1978, 29-30), and, on the other hand, its braking system is out of service. These philosophers believe that as the driver of the train does not have any responsibility for possible movements and events that may occur and does not deserve moral praise or blame, a person who has to do or omit a particular action as a result of divine knowledge or providence or causal determination and cannot choose an alternative has no intentional agency and will not, thereby, be responsible for the action.

According to this traditional view, moral responsibility and even the agent’s personality require him to have more than one alternative, and since many paths should be ahead of the train, he is able to choose from among several options and therefore be responsible for his action.



Principles of Free and Responsible Moral Action

According to the widely held view by Muslim philosophers, free will does not conflict with causality, and even the determination arising from the principle of causation does not necessitate the effect until the will occurs as the final part of the complete cause.[2] Indeed, the denial of causality within the domain of human actions jeopardizes human causality and free will. In order to explain this common viewpoint, we should first consider the implications of the principle of causation:

1. The principle of causal necessitation. This non-constrainable rational principle is formulated as follows: “As long as an object is not necessitated, it will not come into existence” (Ibn Sīnā 2000, 2:548-9; Mullā Ṣadrā 1981, 7:112). According to this principle, insofar as something is not necessitated by a complete cause, it will never come into existence. This requirement is grounded in the quiddity-based contingency (al-imkān al-māhuwī) of the effect, because, as widely believed in Islamic philosophy, “quiddity qua quiddity has no mode of being,” which means that what is necessary for a quiddity is only the predication of its essence and essential properties, and, on its own, the quiddity is neither necessarily existent nor necessarily nonexistent. On its own, it is compatible with both existence and nonexistence; it has essential contingency, so to speak. It is clear that such a quiddity neither can cause its own existence nor can it cause its nonexistence, because both cases imply preponderance without a preponderator (tarjīḥ bi-lā murajjiḥ), which is impossible. Thus, preponderance of the existence or nonexistence of the quiddity requires something else: the complete cause, which necessitates the existence of the quiddity, making it necessary by something else (wājib bi-l-ghayr). As Shaykh al-Ishrāq Suhrawardī believes, if an effect is not necessitated by its complete cause, its realization will be possible, because if it were impossible for it to be realized, it would not come to existence. On the other hand, the realization of an effect is contingent with respect to everything else; that is, it is possible for it to exist in anything, even if it is not a cause; then, there will be no difference between the complete cause and the other things. Thus, a complete cause will no longer be a complete cause, which is contrary to our initial assumption that it was a complete cause (Suhrawardī 1396 Sh, 2:63).

2. The principle of the impossibility of preponderance without a preponderator. According to this principle, which is accepted by many Muslim philosophers and theologians, human free actions require, in addition to efficient causes, purposes and preponderators, in virtue of which they are made unequal with respect to doing and omitting to do actions. In other words, a free agent cannot choose one of the two actions when they are equal in terms of interests, which prompt the agent and serve as his purposes.

Thus, every contingent being is, in its very existence (intrinsic [nafsī] existence) or in the attributes of its existence (attributive [naʿtī] existence), preceded by a cause which grants necessity and determination to it, and as long as a contingent being does not receive this hierarchical (rutbī) necessity through its complete cause, it will never come into existence. Human will, be it equal to a keen appetite and thus passive or a resolution and determination and thus active, as well as the free action following the will are necessitated by prior causes and ultimately by the Necessary Being. However, this is not in conflict with human free will. From a philosophical viewpoint, power and free will mean that “an agent is such that he or she can do a particular action if he or she wants to, and does not do it if he or she does not want to” (Mullā Ṣadrā 1981, 6:307). This conditional proposition has no implications for the external impossibility or necessity of its referent and only states the possibility of the realization of a free action and the power of the willing agent to do the action. Now, if a person does an action based on his or her will and that contingent action becomes necessitated, the necessity resulting from the agent’s will does not contradict the agent’s free will more than it confirms it.[3] This is because the will in question is part of the action’s complete cause.

However, the question arising from the definition of free will proposed by Muslim philosophers and the role of the agent’s will therein is whether the will, as the last constitutive of human action, is itself voluntary and free or not. Does this free will itself depend on another will and so on, leading to a regress, or is there finally a point at which human will is necessitated without the agent having any dominion over action or omission (determinism)? Some Muslim philosophers have explicitly accepted determinism. For instance, Mullā Ṣadrā considers cognition, appetite, and will as the roots of human free action and parts of a necessary causal chain that starts with the Necessary Being such that human will and action are necessitated by the divine essence, and the agent has to do the action if the divine essence requires so, and if the divine essence does not thus require, the agent will not have the ability to do the action. Thus, although human free action is dependent on his will, this will is necessarily and deterministically produced such that the agent plays no role therein.

Mullā Ṣadrā’s argument for the deterministic nature of the will, following al-Fārābī and Mīr Dāmād, is that if the will were voluntary, it would be the effect of another will, and this will, if voluntary, would in turn be the effect of another will, and so on, leading to a regress. Thus, although human external action is free and dependent on his will, the will itself is not free or based on a further will. Instead, it is deterministic and involuntary (Mullā Ṣadrā 1981, 6:390; Fārābī n.d., 73-4; Mīr Dāmād 1995, 445; Muḥaqqiq Dawānī 2002, 155).

On this conception, human agency is completely involuntary like the agency of the fire, because both agencies are affected by a series of external causes, and as fire cannot avoid burning, a person who does a particular action cannot avoid the action, because his action and will have certain causes beyond his or her free will, and when these causes take place, the realization of the effect will be necessary. The only difference between human agency and natural phenomena like fire is human knowledge and awareness of the promptings and purposes that compel him or her to a particular will and action.

Thus, in the common philosophical definition of free will, the free agent is reduced to a willing agent whose will is preferred and necessitated as a result of external causes,[4] without him or her having the ability to will or not to will or, for that matter, to do or not to do an action.

Given this problem, Some Muslim thinkers and scholars of the principles of jurisprudence (uṣūl al-fiqh), such as Muḥammad Bāqir Waḥīd Bihbahānī, accepted the definition of free will proposed by theologians, fully denying the necessity between causes and effects (Bihbahānī 1415 AH, 385). Some scholars, like Ākhūnd Khurāsānī, believe that only some of the constituents of free will are free (Khurasānī 1407 AH, 124) and some others, such as Mīrzā Mahdī Iṣfahānī, Muḥammad Ḥusayn Nāʾīnī, Abu l-Qāsim Khūʾī, and Muḥammad Bāqir Ṣadr, try to save human free actions by making a differentiation in this principle. These scholars believe that the principle of causal necessitation is false and impossible at least for these agents, because of the contradiction with the agent’s free will. Thus, they believe that the realization of an effect when its complete cause occurs is neither preferable nor necessary. Now, since the widely held view of philosophers requires determinism, we should either yield to their causal principle, which implies prior necessitation and determinism, or reject their view because of the intuitive and inevitable fact of free will, because the will and its roots (conception, judgment, inclination, appetite, and the other facrots involved) are not under the control of the soul and result instead from involuntary factors. In cases where the soul deliberates on something’s advantages and disadvantages and produces a conception, a judgment of benefit or advantage, appetite, and will, the emergence of psychological desire is imposed on the soul entirely from outside (cf. Iṣfahānī 2008, 405-10; Khū’ī 1419 AH, 1:134-41; Fayyāḍ 1996, 2:42-66; Ṣadr 1417 AH, 2:27-39).

Evaluation of the View of Muslim Philosophers

The question is how defendable free will is based on PAP and its necessary and deterministic principles in the view of Muslim philosophers. Before providing an answer to this question, it should be noted that, in his important and influential paper Alternative Possibilities and Moral Responsibility, Harry Frankfurt challenges the common notion of freedom in PAP for moral responsibility. He shows that although in cases of hypnotism and internal compulsion
we sometimes encounter conditions in which the agent is both forced
to do a particular action and it is totally impossible for him or her to
act otherwise—and therefore according to PAP, his or her
moral responsibility will be meaningless—we should nonetheless note that:

FP: In some cases, there are conditions that do not actually impel the person to do a particular action but in which it is impossible for him or her to avoid doing it. However, the agent will be considered in these conditions as morally responsible. (Frankfurt 1969, 830)

FP provides an opportunity for philosophers who want to endorse both the causal determination and the moral responsibility of the agent. In this recent semi-compatibilism, moral responsibility is compatible with causal determination, whether free will is identified with the ability to act otherwise or not, because moral responsibility does not basically depend on the agent’s ability to act differently than what he or she has actually done.

Consider the Frankfurt-type example of the “unlucky mayor,” which was cited by Fischer and Raviza in 1991 in a paper titled Responsibility and Unavoidability. We review it in this section with some modifications:

Assume Babak is very dissatisfied with the tax programs of the mayor, and because he thinks there is no way to convince the mayor, he tries to kill him. Although these reasons do not seem justified, they are convincing for Babak. Babak is not hypnotized or brainwashed, and he has not been under pressure by another person. He thinks fully reflectively and devises a plan for the mayor’s murder. Furthermore, Babak tells his reliable friend Ahmed about his plan. Babak is an evil person but Ahmed is worse, because he has a personal spite towards the mayor and is completely satisfied with Babak’s plan. What is more, he considers a more dangerous plan. He worries that Babak may waver from his plan, so he secretly plants a device in Babak’s brain so that he can monitor and control all Babak’s brain activities and intervene if necessary. Ahmed wants to guarantee the mayor’s murder if Babak changed his mind. Anyway, Babak and Ahmed both go to a meeting at the townhall and Babak kills the mayor according to his own plan, and consequently Ahmed does not intervene in the murder (Fischer and Ravizza 1991, 258-9).

According to the Frankfurtian assessment in this case, it is obvious that Babak is morally responsible for the mayor’s murder and deserves the blame even though he did not have any other possible alternative and was not able to avoid the murder (because if he was dissuaded from the murder, the device planted in his brain by Ahmed would make him kill the mayor). Thus, in cases in which a responsibility-undermining factor operates in the alternative sequence, while the actual sequence of the action is not thereby affected, the agent will be held morally responsible for his action even though he or she could not have acted otherwise.

Now, we return to the question of this section regarding the view of Muslim philosophers regarding free will. Incompatibilists like Van Inwagen and Robert Kane require two basic conditions for human moral free will and responsibility. The first is that possible alternatives are present for the moral agent, from which he or she is free to choose; the second is that the origin of the agent’s actions lies in him or her and not in something outside the domain of the agent’s will (see Van Inwagen 1975, 185-99; Kane 2005, 120-31). Considering the fact that Muslim philosophers consider the agent’s will or free will as the final part of the action’s complete cause and the determination and fulfillment of free action basically depends on this element, not only is there no contradiction between the agent’s free will and the causal determination of his action but also these two have been considered as concomitant in every free action. Thus, causal determination is not an external reality independent of free will, and so it never contradicts it. However, when we think about the common definition of free will offered by Muslim philosophers (i.e., an agent being such that he or she can do a particular action if he or she wants to, and does not do it if he or she does not want to [Mīr Dāmād 2002, 1:94; Ibn Sīnā 1404 AH, 173]), it becomes clear that it is, on the one hand, on a par with PAP and subject to Frankfurt’s objection and, on the other hand, it suffers from the absence of the second condition introduced by incompatibilists—namely, that the agent of an action has free will to do that action only as long as he or she was able to avoid the action in a counterfactual case in which he decided not to do the action. In contrast, according to Frankfurt-type cases, the free will and responsibility of moral agents do not depend on alternative possibility and the ability to have avoided doing what one has actually done. Consider the following two examples for the assessment of this claim. First, we review the “accident 1” example:

Ahmed starts his car and drives it. A criminal gang, unbeknownst to him, have tampered with his car in a way that he will not be able to stop his car or change its direction. Ahmed runs into a pedestrian on his way. As it happens, the pedestrian is someone who is strongly hated by Ahmed. Ahmed hits the pedestrian and kills him, without any deliberation and decision to change directions, reduce speed, or stop. 

As can be seen in this example, had Ahmed intended an alternative action, he actually would not have been able to do so. However, he could decide to prevent the accident.

In the “accident 2” example, in addition to tampering with Ahmed’s car, the criminal gang plant a device by a conversant neurosurgeon in his car beside his head such that his cerebral waves are completely controllable. The neurosurgeon will intervene to bring about a decision to hit the pedestrian if Ahmed shows an inclination to decide to refrain from it. Of course, Ahmed makes his own decision based on a prior spite and hits the pedestrian. In this example, there is no external alternative for Ahmed and he has no power to decide to do an alternative action. He cannot even make a decision to prevent the accident.

Now, in the “accident 3” example, Ahmed is not able to stop his car, reduce speed, or change directions as a result of a deliberate tampering and he does not have any brain control, and so Ahmed’s choices and decisions are under the control of the criminal gang. In addition, in this case, he becomes aware of the plot just a few seconds before the accident via one of his friends who is a gangster. He finds out that he cannot make a decision other than the gang’s decision and cannot prevent the accident. So, without any thoughts about any moral principles, he attempts at making a good decision, passively continues to drive, and ultimately crashes into the pedestrian.

As can be seen, contrary to the view of Muslim philosophers and PAP on free will and moral responsibility, in all the three cases, the agent lacks the ability to avoid his action. But should we regard the agent as not responsible for the accident in all the three cases? Our moral intuition is that, contrary to the view of Muslim philosophers and PAP, in the first and second examples, despite the fact that the moral agent does not have any practical alternative and is not able to change his surroundings, he is morally responsible and deserves blame. In the “accident 1” example, he was able to make a decision to prevent the accident, but he did not even begin to make the relevant choice or decision with complete cruelty. Thus, counterfactually speaking, if the car had not been defective, this accident would still have happened. In other words, the car’s technical defect did not have a real effect on the agent’s action. In the “accident 2” example, although Ahmed neither could change the speed or direction of his car nor could make such a decision, he still believed that he had the power for deliberation, choice-making, and decision-making, but he finally hit the pedestrian with his own will and is, therefore, responsible for his action.

However, by comparing example 2 with example 3, the problem with a Frankfurt-type view, according to which the presence of an alternative does not have any role in moral responsibility as well as the elliptical problem in the definition of free will in Islamic philosophy, which may restrict the consequent of “if he or she did not want it, he or she would not do it” to mental actions such as the agent’s decision-making, becomes obvious. Although an external and concrete alternative (as in example 1) and even an internal alternative (as in example 2) are not necessary for the realization of the free will and moral responsibility of the agent, it is undoubtedly necessary for the agent to think that such alternatives are accessible in counterfactual circumstances and that he is able to decide and act differently if he so wants.[5] Therefore, in the third example where the agent has no alternative and is aware of such conditions, he cannot be deemed responsible for the events that happen.

But the second condition of incompatibilists for the responsibility of the agent was the restriction of the origin of the agent’s actions to herself, as she is ultimately responsible for the actin. As Frankfurt suggests in his paper Freedom of the Will and the Concept of a Person (1971, 5-20), it is necessary to distinguish between freedom of action and freedom of the will in the discussion on moral responsibility. A person’s freedom of action means that he can freely go for any action: he can talk, walk, read a book, and so forth, but freedom of the will refers to the freedom of the agent in everything he wants. In other words, apart from the fact that a person is free in his or her action and can, for instance, walk freely, he is also free in the second-order will, which has this will its object.[6]   

Following Frankfurt, philosophers like Watson have welcomed the idea of the levels of desire and will (Watson 1975, 217-8). They unanimously suggest that human moral responsibilities do not purely depend on the fact that one’s actions are under the control of his will, but the responsibility is in virtue of the fact that the person’s will is not a pure psychological characteristic and does not result from an external factor, but it manifests some features which are originated from the agent. Now, in the view of Muslim philosophers on free action and its perceptual and emotional roots, from the time when a simple conception takes shape till we come to keenness in appetite, desire, decision, and the will for doing an action, to what extent can we consider the agent the final source and responsible for these roots and ultimately for the voluntary action? Do the resulting will and its prior contingent bases and causes depend on another desire and will within the agent? Does this type of voluntarism proposed by Muslim philosophers not seem to require another will?

Muslim philosophers have been aware of this problem (cf. Mīr Dāmād 2002, 209; Mullā Ṣadrā 1981, 4:114; 2002, 1:207; Javadi Amoli 2007, 2-4, 75) and have suggested that such an intentional agent (al-fāʿil bi-l-qaṣd) will be constrained in his or her action and thus different from other kinds of agents such as the agent by providence (al-fāʿil bi-l-ʿināyah) and the agent by agreement (al-fāʿil bi-l-riḍā). On this model, an agent’s will is not identical to her essence or quiddity, and as a created thing, it requires a creator, which is the agent’s own will or something else. It is obvious that the first option leads to a regress of wills. Thus, we should accept the second option and truly reduce the agent’s will to an involuntary phenomenon dependent on something else. Surely this will is a coercive fact. This is because, on this conception, the agent’s existence and actuality depend on the will of an external entity.

In order to deal with this problem, some philosophers try to introduce a mental and internal event in the agent as causing his free actions, which is not itself a kind of personal action (see Hornsby 1986, 275-86). But, instead of travelling such an uneven path and attempting to justify a mental phenomenon that, on the one hand, is not a free action, and its annexation to human action, on the other hand, is beyond the attribution of some psychological or mental affections and makes the action “free,” we can in the first step contemplate the solution that free will and moral responsibility are gradational and hierarchical realities. Thus, although human beings do not have absolute and complete responsibility for their voluntary actions, as well as for all of its prior causes and effects, our moral intuition differentiates between an intentional agent who does a blameworthy action like murder with his own first-order desire and choice and a completely coerced agent who lacks this primary desire and will and freedom of action. Moral responsibility of individuals is commensurable with this level of the freedom of will. The agent might be deemed responsible as a result of freedom of action and his will for an action, while he might not be responsible for an action at a higher level, because he did not have any will for this will. Thus, lack of complete responsibility on the part of an individual is not concomitant with lack of partial and restricted responsibility.

It is clear that despite the agreement on commonsensical intuitions, such an answer is still subject to an elliptical problem of incompatibilists that the reduction of will to an inevitable product of certain deterministic causes—or the lack of a will for the will—amounts to the lack of will in action and hence the absence of responsibility, because  there is indeed no real and effective difference between an agent who does not have any will in his action and a person who has a will that has been imposed on him unwantedly and involuntarily—the latter has the will to lie, observe traffic rules, and so on, but this will is also part of a process similar to the processes that lead to a robot’s action. He inevitably has the will to do an action, and cannot preclude the will.

In the face of this problem, let us note that a robot that deliberates and does an action finally with his own will is different from a robot that is completely submissive on his maker like a puppet and has no opportunity to think and choose.[7] Furthermore, it should be noted that an agent’s will is by no means a linear event, which finally leads to a voluntary or non-voluntary phenomenon; rather, it is formed in a complex network of desires, beliefs, and wills, and along with external factors, and by any change at each stage of the chains of this network, transformations may happen in the final result. In this integrated network where very simple and primary desires and wills to higher human wills and desires are found, the realization and continuity of prior causes that lead to a specific will are basically voluntary and free. Thus, not only is it possible for the agent to provide the grounds for many conceptions, judgments, and primary appetites, but also he can prevent prior judgments and tendencies at each stage and thus the formation of a particular will by considering alternative beliefs and desires.


In this paper, I first modified the definition offered by Muslim philosophers as well as the Frankfort-type model in which free will is believed to hold when an agent thinks he or she has an alternative possibility when choosing and doing an action. Therefore, he would be able to prevent the action in counterfactual circumstances. In the next stage, I pointed out that although deterministic principles of human will take away his full-fledged and absolute free will, contrary to what some Muslim philosophers and scholars of the principles of jurisprudence have suggested, the agent cannot be considered as coerced or determined, because some levels of free will can still be found in the agent, which differentiates him from creatures lacking that level of free will.

[1]. Michael Otsuka defends a different but related principle called the “principle of avoidable blame” (PAB):

PAB: One is blameworthy for performing an act of a given type only if one could instead have behaved in a manner for which one would have been entirely blameless (see Otsuka 1998, 688).

He believes that PAB is plausible, requires incompatibilism, and has more stability compared to PAP.

[2]. According to a widespread tendency in Islamic philosophy, in the process of doing or not doing an action, the action comes to the agent’s mind first, and the agent imagines the goodness of doing or omitting the action. When the benefits of the action are affirmed by the agent, an appetite is formed, and with the intensification of the appetite, a resolution or firm appetite is formed for doing or not doing the action, which is called will (c.f., e.g., Mulla Sadra 2002, 1:207).

[3]. In philosophical terms, the necessitation resulting from free will does not conflict with free will.

[4]. For instance, in his discussion of psychic qualities and of the principles of the human action, ʿAllāmah Ṭabāṭabāʾī suggests that there is a completely necessary relationship among these principles such that after the formation of a firm appetite, the will emerges necessarily (Ṭabāṭabā’ī 2008, 122).

[5]. Thus, an insolent (mutajarrī) agent who has, due to his accumulated ignorance, considered the proposition “A is B” true and lied nevertheless by saying that “A is C,” while A was in fact C, has done something wrong and is morally responsible for it, because he has had an immoral intention and was able to avoid it. What is more, even if it was impossible for him to avoid the immoral intention and he was ignorant of the flaw but yet decided to lie, then he is still responsible for his insolence (tajarrī).

[6]. Frankfurt makes use of the phrases “first-order desires” and “second-order desires” for explaining this distinction. First-order desires are desires for actions, for achieving goals, or avoiding certain circumstances, while second-order desires are desires for the existence of certain first-order desires. Frankfurt believes that if an agent wants to have both freedom of action and freedom of the will, then in addition to having first-order desires, she should control her first-order desires with second-order ones. However, it seems that  freedom of will ipso facto requires both desires.

[7]. It is clear that as internal and external restrictions governing a given agent increase, the domain of his or her will and free will will be further restricted correspondingly. But firstly the legal and moral responsibility of an individual is proportionate to the domain of his or her free will, and secondly the material agent, whether a human or a robot, cannot evade these restrictions. A free agent like human beings can decide freely and based on his or her choice and decision. For example, he or she can go from Tehran to Mashhad in his car or stay at home and study behind his personal computer. However, he will be free only in certain frameworks, because he cannot will his internal wills and cannot truly determine their occurrence; he cannot choose a direct route to go to Mashhad instead of choosing bendy roads; and if he stayed at home, he had a freedom of action only within the range of predetermined computer programs. Anyway, he will be responsible within the range of his freedom. So, we differentiate between this person and a person who lacks even this level of free will.


Bihbahānī, Waḥīd. 1415 AH. Al-Fawāʿid al-ḥāʾiriyyah. Qom: Majmaʿ al-Fikr Al-Islami.

Farābī, Muḥammad ibn Muḥammad. n.d. Fuṣūṣ al-ḥikam. Tehran: Bidar.

Fayyāḍ, Muḥammad Isḥāq. 2006. Muḥāḍarāt fī uṣūl al-fiqh. Qom: Ansarian.

Fischer, John Martin, and Mark Ravizza. 1991. “Responsibility and Inevitability.” Ethics 101 (2): 258-78.

Frankfurt, Harry. 1969. “Alternate Possibilities and Moral Responsibilities.” Journal of Philosophy 66 (23): 829-39.

Frankfurt, Harry. 1971. “Freedom of the will and the concept of a person,” Journal of Philosophy 68: 5-20.

Goldman, Alvin I., and Kim Jaegwon, eds. 1978. Values and Morals: Essays in Honor of William Frankena, Charles Stevenson and Richard Brandt. Dordrecht: D. Reidel.

Hornsby, Jennifer. 1986. “Bodily Movements, Actions, andMental Epistemology.” Midwest Studies in Philosophy 10: 275-86.

Ibn Sīnā, Ḥusayn ibn ‘Abd Allāh. 1404 AH. Al-Shifāʾ (al-Ilāhiyyāt). Qom: Ayatollah Marʿashi Najafi Library

Ibn Sīnā, Ḥusayn ibn ‘Abd Allāh.  2000. Al-Najāt. Tehran: Tehran University Press.

Isfahānī, Mīrzā Mahdī. 2008. Abwāb al-hudā. Edited by Husayn Mufīd, Tehran: Monir Cultural and Publication Center.

Javadi Amoli, Abdollah. 2007. Raḥīq makhtūm. Annotated by Hamid Parsania. Qom: Esra Publication Center.

Kane, Robert. 2005. A Contemporary Introduction to Free Will. New York: Oxford University Press.

Khuʾī, Abu l-Qāsim. 1419 AH. Ajwad al-taqrīrāt. Qom: Sahib al-Amr Institute.

Khurāsānī, Muḥammad Kāẓim. 1407 AH. Fawāʾid al-uṣūl. Tehran: Ministry of Islamic Guidance and Culture.

Levy, Neil, and Michael McKenna. 2009. “Recent Work on Free Will and Moral Responsibility.” Philosophy Compass 4 (1): 96:133.

Mīr Dāmād, Muḥammad Bāqir. 2002. Collected Works. Tehran: Association of Cultural Works.

Mīr Dāmād, Muḥammad Bāqir. 2005. Qabasāt. Tehran: Tehran University Press.

Muḥaqqiq Dawānī, Mullā Ismāʿīl. 2002. Sabʿ rasāʾil. Tehran: Mirase-Maktoub.

Mullā Ṣadrā, Muḥammad bin Ibrāhīm. 1981. Al-Ḥikmat al-mutaʿālīyyah fī l-asfār al-ʿaqlīyyah al-’arbaʿa. Beirut: Dar Ihyaʾ al-Turath al-Arabi.

Mullā Ṣadrā, Muḥammad bin Ibrāhīm. 2002. Al-mabda’ wa l-ma’ād. Edited by Muhammad Zabihi and Jafar Shah Nazari. Tehran: Sadra Islamic Philosophy Research Institute.

Otsuka, Michael. 1998. “Incompatibilism and the Avoidability of Blame.” Ethics 108: 685-701.

Qūshchī, Alī bin Muhammad. n.d. Sharḥ Tajrīd al-ʿaqāʾid. Qom: Razi and Bidar.

Rāzī, Fakhr al-Dīn. 1986. Al-Arbaʿīn fī uṣūl al-dīn. Cairo: al-Azhar University Press.

Rāzī, Fakhr al-Dīn. 1411 AH. Al-Mabāḥith al-mashriqiyyah fī ʿilm al-ilāhiyyāt wa-l-ṭabīʿiyyāt. Qom: Bidar.

Ṣadr, Muḥammad Bāqir. 1417 AH. Discussions on Jurisprudence. Qom: Islamic Jurisprudence Encyclopedia Institute.

Suhrawardī, Shahāb al-Dīn Yaḥyā, 1396 AH. Collected Works. Edited by Henry Carbon. Tehran: Iranian Institute of Philosophy.

Ṭabāṭabāʾī, Muḥammad Ḥusayn. 2008. Nihāyat al-ḥikma. Qom: Boostan-e-Ketab.

Taftāzānī, Masʿūd ibn ʿUmar. 1409 AH. Sharḥ al-Maqāṣid. Edited by Abd al-Rahman Amirah. Qom: Manshourat al-Razi.

 Van Inwagen, Peter. 1975. “The Incompatibility of Free Will and Determinism.” Philosophical Studies 27: 185-99.

 Watson, Gary. “Free Agency.” The Journal of Philosophy 72: 205-20.