Contemporary philosophers usually approach the notion of freedom by the distinction between freedom of choice and freedom of action. The implicit assumption is that, although freedom concerns both willing and acting, the kind of freedom involved in volitions is an essentially different thing from the kind of freedom involved in actions. Given this difference in kind, it seems there is no need to treat the second in relation to the first. So, while some (few) philosophers traditionally focus their researches on the freedom of choice as the necessary condition for free acting, most contemporary theorists deal with freedom exclusively from the viewpoint of its possible explanatory role in agency (believing the attribution of freedom to human subjects substantially independent on the possession of a faculty as free will).

The debate on freedom admits four basic views: 1) Libertarianism; 2) Hard Determinism; 3) Compatibilism (or Soft Determinism); 4) Skepticism. Since I simply aims at briefly sketching the categories in the contemporary taxonomy of positions, I will not provide a detailed account for each of the views. Particularly, I will use a disjunctive formula in order to capture both those philosophers focusing on freedom of choice and those focusing on freedom of action.

The libertarian theorist holds that a human subject is free if and only if either the subject’s choices or the subject’s actions are ultimately up to her.¹ This dependance of either choices or actions on the subject is characterized by two assumptions:

a) the subject is the source of her choices or the subject is the source of her actions;

b) the subject is able to choose or act otherwise than what she actually does.

Since the libertarian believes that she can make a case for the soundness of “(a) & (b)” as the right understanding of human willing or agency, the libertarian then adfirms that human subjects possess freedom.

Contrary to this view, the hard determinist theorist holds that “(a) & (b)” fails to grasp the story about human freedom; since being determinism true, then (b) can not be the case. Indeed, given the laws of nature and a set of initial conditions, everything should be necessitated to occur by the preceding conditions (of the universe): therefore, neither the subject’s choices nor the subject’s actions can ever be up to her.² In reason of this, Libertarianism and Hard Determinism reaveal to be incompatible positions: if Libertarianism is true then Hard Determinism is false; if Hard Determinism is true then Libertarianism is false.

Compatibilism challenges both libertarianism and hard determinism claiming that, since ordinary attributions of freedom to human subjects do not originate from considerations on the metaphysical frame of nature (being rather a matter of moral concern), the assumption of determinism does not prevent that human subjects can consistently be understood as equipped with free will or as free agents.³ The traditional compatibilist idea is that in order to understand freedom it simply suffices that (a) is true, leaving aside the validity of (b).⁴ According to compatibilism common sense attitude to moral responsibility confirms this idea. Indeed, common sense reactions to morally relevant actions swing from blame to praise. This entails that ordinary attitudes toward these notions make sense just within a framework in which human beings are thought to be morally responsible for their actions. But to be morally responsible seems to imply that the subject of moral responsibility is free, that is, she has some kind of control over his actions. This kind of freedom is then simply understood in terms of responsibility for actions or control over these, independently on the possibility to do otherwise.

Finally, the skeptic theorists holds that, independently on which of the established views is true, neither of these actually grasps what freedom is. The skeptic point of departure is that “(a) & (b)” expresses the common sense intuition of freedom. By this assumption all theories result faulty. Suppose Libertarianism is true: if this is the case, given (b), there would be no costraining reasons for the subject’s choices or actions. Therefore, human subjects should appear determined to will or act by chanche rather than by freedom. On the contrary, suppose Hard Determinism is true. Then human subjects reveal to be not free. Finally, suppose Compatibilism is true: the common sense attitude to the assumption of (b) is rejected. Then Compatibilism reveals to be problematic: for a side the theory makes appeal to the common sense attitude towards moral responsibility, for the other the theory rejects the common sense intuition of freedom.

Now, although the contemporary debate appears very useful in order to clarify the notion of freedom and the metaphysical difficulties it involves, my position is that the problem of freedom resists such an assessment for two main substantive reasons (that is, reasons that get their explanatory value by grasping metaphysical features of freedom).

Firstly, the distinction between freedom of choice and freedom of action is at best pointless, at worst senseless: prima facie a subject is free if and only she is metaphysically free to choose (as she likes); that is, if her choices are ultimately up to her (I do not mean that allchoices of a subject should be up to her, but that at least the most important ones should be up to her). Indeed, suppose a human subject someway free to act but not free to will. Would anyone actually consider this subject free? I fear that the answer is no, since there is no intellegible intuition by which the expression free to act without being free to willmakes sense when predicated of a human subject. On the contrary, suppose a subject has no chance to choose among different options: her behaviour is completely determined by the state of affairs she is experiencing. Moreover, suppose that subject has decided with no costraints to conform her will to the only action she can perform (for example, God appears to me and command me you shall follow my divine decrees from now on). Would anyone actually consider the subject not free in behaving as the situation determine her to act? I believe the answer is noagain. Consider my example: if the being appearing to me is really God, I have no possibilities to resist Him (probably I could not face my fear, my anguish, my delight, my wonder); nonetheless, I would be free in my accepting His command (that is, my freedom would result happily conforming my will to His, even if I probably have not possibility to do otherwise). In reason of these considerations, I conclude that free agency is commonly understood simply as a practical consequence of a free volition: if an action freely flows from a free volition, than the action is free; if an action freely flows from a volitions that is not free, than the action is not free too.

Secondly, the four views advanced by contemporary theorists do not cover all the theoretical positions concerning the notion of freedom. That’s why: almost all contemporary theories move from a common characterization of the volition-forming process or the action-production process that do not grasp all the actual formations of volitions or the productions of actions. Consequently, those theories appealing to a characterization of volition-forming process or action-production process alternative to that endorsed by the positions (1), (2), (3), and (4), necessarily fall outside the set of theories on freedom defined by the union of positions (1), (2), (3), and (4). A very short justification for my claim. The common characterization of volition-forming process and action-production process can be spelt out by the conjunction of the following assumptions:

c) The volition-forming process or the action-production process are to be referred only to the subject either of the volition or of the action (at least for those volitions or actions which are relevant for the attribution of freedom to the subject);

d) The volition-forming process or the action-production process are linear processes; that is, processes where different reasons or causes count by their absolute weight or force in occurring either the volition or the action;

e) The volition-forming process or the action-production process occur istantaneously.

Given the phenomenology of decision making, by which habits, aptitudes and inclinations often determine a subject to will or act in a certain way more unavoidably than the most forcefull reasons can, assumptions (c), (d), and (e) appear really problematic to hold. Volitions and actions don’t seem to be always in control of the subject; nor the processes forming the ones or producing the others appear to be linear and instantaneous.

Therefore, “(c) & (d) & (e)” does not provide the only possible characterization of volitions and actions. My point is not to dismantle the adequacy of “(c) & (d) & (e)”: it is simply to reject the claim that this conjunction of assumptions covers all types of volition-forming process and action-production process theories.

Consequently, if “(c) & (d) & (e)” reveals to be one among other characterizations for the volition-forming process and the action-production process, it is natural to expect that theories not endorsing “(c) & (d) & (e)” will not fall under the set of theories of freedom defined by the union of (1), (2), (3), and (4).

Now, my purpose in this paper is to provide reasons against the contemporary approach on freedom, in order to show that a more promising account than one or the other version of the positions (1), (2), (3), and (4) is necessary. Particularly, I will try to highlight that (C₁) the distinction between freedom of choice and freedom of action does not grasp any important feature of freedom (freedom being understood prima facie as freedom of choice) and that (C₂) at least a relevant theory not endorsing “(c) & (d) & (e)” is philosophically available. I name this account of freedom, which is the one I favour, Theory of Procedural Agency (TPA).

Although a positive defense of (C₁) and (C₂) plus ad hoc arguments against (1), (2), (3), and (4) is the most systematic way to proceed, I will go along a different path. The reason is simple: I construe my paper as a preliminary work, mainly directed to dismantle the air of completeness and self-confidence that the contemporary taxonomy of views on the metaphysics of freedom seems to have. To this end providing a case which does not fit the taxonomy better supply my intention as to (C₂) than any other method. And since this case offers good reasons both in support of (C₁) and against the soundness of “(c) & (d) & (e)”, I believe my case to accomplish my preliminary task.

The case I argue for are Berkeley’s considerations about free will and agency. My idea is that these considerations furnish an embryonic approach to TPA. This does not mean that Berkeley has developed a consistent and complete theory: I simply claim that his treatment of the topic provides the theorist with very insighful intuitions and arguments in order to rightly setting the notion of freedom. My paper is an endeavour to work out these intuitions and arguments into a preliminary version of TPA.

In the first section of my paper I will present the context within which Berkeley treats the notion of freedom. I will consider just arguments from Alciphron: this work is not addressed only to professional philosophers, rather aiming at a larger audience. For this reason the book presents philosophical topics as plain dialogues among learned men. As a consequence Berkeley takes pain to draw a general picture of the philosophical debate in his era. This picture shows adequately that, as to the topic of freedom, the philosophical debate in the eighteenth century was deeply in line with the traditional assumption of (C₁). In the second section of my paper I will raise some criticisms against Berkeley’s theory of free will. Particularly, I will show that in order to assume consistenly (C₁) there’s a need for a distinction among different kind of volitions. To this regard I claim that the assumption of a constitutive view of self knowledge would help. In section three I will explain Berkeley’s claims concerning agency; and I will provide a logical formulation of these as an embryonic version of TPA. I will show that the theory is a case in support of (C₂) and that, given that, there are good reasons to reject “(c) & (d) & (e)”. Finally, in the last section, I will show some of the difficulties a berkeleian like TPA involves, claiming that a sound version of TPA should answer positively these difficulties.

1.

In the seventh dialogue from Alciphron Berkeley deals at length with the topic of free will. The context of the discussion is the rationality of religion. The advocate of freethinking defends the claim that religion is “an unreasonable absurd thing” (Alc., VII. 19, p. 258).⁵ The claim is formally the conclusion of a sequence of propositions:

Religion implies the worship of a God, which worship supposes rewards and punishment, which suppose merits and demerits, actions good and evil, and these suppose human liberty, a thing impossible: and consequently religion, a thing built thereon must be an unreasonable absurd thing (Alc., VII. 19, p. 258).

Now, despite appearances, the sequence of propositions does not represent a sound argument: although Alciphron lists them stating they relate the one to the other by supposition, this relationship appears to be mostly unclear. Nonetheless, I believe the sequence can be worked out into logical form, which makes explicit what remains mostly implicit. In order to accomplish this task, the freethinker needs principles and premises. I think a set of one principle and three premises can do the work:

Principle: Religion implies the worship of a God.

Premise₁: The worship of a God consists of actions performed in conformity with the content of religious beliefs.

Premise₂: A creature truely worshipping a God achieves a reward by God.

Premise₃: A creature worshipping a God deserves reward by God for her worship iff she can choose between acting in conformity or not with the content of her religious beliefs.

Berkeley’s spokesmen endorse the principle and the three premises. As to the Principle: at the end of the fourth dialogue Alciphron criticisizes Christianity presuming that religion and divine worship are the same thing (identity thesis). Crito answers to his criticism without rejecting this identity thesis (Alc., IV. 25, pp. 140-141). Furthermore, the sequence of propositions against the rationality of religion explicitly begins with the assumption of the Principle. Euphranor’s answer, as Crito’s, does not reject the identity thesis.

As to Premise₁: Berkeley devotes the fifth dialogue of his book to argue for the claim that Christian faith induces mankind to practice virtues as charity or justice in accordance with the content of Christian principles. For example, at paragraph 14 Crito defends this view comparing Roman and British history. He says:

an indifferent eye may… perceive a vein of charity and justice, the effect of Christian principles, run through the latter [the British history]; which, though not equally discernible in all parts, yet discloses itself sufficiently to make a wide difference upon the whole, in spite of the general appetites and passions of human nature, as well as of the particular hardness and roughness of the block (Alc., V. 14, p. 154).

As to Premise₂: Berkeley believes that natural religion, by the simple use of reason, adequately grasps the doctrines of Providence, of the immortality of the soul, and of a future state of rewards and punishments (Alc., V. 27, p. 169). Nonetheless, men are rarely “swayed or governed by mere ratiocination”, being more aptly lead by faith in most affairs of life.⁶ Consequently, the revelation by God communicate saving truths in a manner appropriate to mankind (Alc., VI. 8, pp. 194-195; Alc., VI. 31, p. 232), i. e. as instructions for worship (Alc., VII. 11, p. 249). Christian worship drives then men to their eternal salvation, that is, Christian worship is rewarded by God.

Finally, Premise₃: at paragraphs 10 and 11 of the third dialogue, Berkeley draws the image of a rational society of agents. According to the supposition, this society is like a city whose founder and monarch is God. Every agent conforms her will to the divine will, concurring to the promotion of the general well-being of the whole. In order to enforce rational agents to obey the established laws, God provides incentives for good actions and discouragements for evil actions by stipulating a system of rewards and punishments.⁷ Now, all this makes sense only if premise3 is assumed. Indeed, first, if God needs to enforce rational agents to obey the law, they evidently could do otherwise; second, if God provides incentives (and discouragements) to good (and evil) actions, God rewards those agents that, having the possibility to perform evil actions, have nonetheless chosen to perform good ones (the contrary for the case of punishing).

From the assumption of the principle and the three premises, the argument follows. The first step infers the consequences from the three premises concerning the concept of worship:

I. a) From Premise₁ and Premise₂ follow that I. (1) a creature acting in conformity with the content of her religious beliefs achieves a reward by God;

I. b) from I. (1) and Premise₃ follow that I. (2) a creature acting in conformity with her religious belief deserves a reward by God for her actions iff she can choose between acting in conformity or not with the content of her religious belief;

I. c) Now, be A an action. It seems evident that if a creature can choose between performing A or not performing A, that creature, having chosen to perform A, could have done otherwise;

I. d) Consequently, from I. (2) and I. (c) follows that I. (3) a creature acting in conformity with her religious belief deserves a reward by God for her actions iff she could do otherwise than what she actually does.

At this stage the freethinker moves to the second step of the argument, raising a case against the notion of a creature having the possibility to do otherwise than what she actually does.

I. e) Since all in Nature is necessary determined, I. (4) no creature has the possibility to act in a different way from how she actually does.

I. f) Then, the concept of worship implies a puzzle. Indeed, if I. (4) is true, the condition required for deserving a reward from God, stated by I. (3), is impossible to satisfy.

I. g) Finally, consider the Principle: from Religion implies the worship of a God and the three premises follow that Religion implies that a creature acting in conformity with her religious belief deserves a reward by God for her actions iff she could act in a different way from how she actually acts. Therefore, if I. (4) is true, it follows that Religion implies that, in order to deserve rewards, it is necessary the satisfaction of a condition that is not possible to satisfy. Since requiring to satisfy a condition, in order to deserve rewards, that is not possible to satisfy appears to be a good candidate for the designatum of the expression “an unreasonable absurd thing”, it seems to me that Alciphron’s claim on religion is logically justified if I. (4) is true.

Now, in order to defend the argument there’s a need for an understanding of I. (e), particularly the expression necessary determined occurring in the premise to I. (4). I will reason in terms of hypotetical freedom and categorical freedom.

Suppose that while I’m writing my paper at time tn, I think to move to my kitchen and take a glass of water. Clearly this is a kind of thing I’m able to do. Suppose further that H is a proposition about the complete state of the world at time t_i (t_i precedes t_n); and Lis a proposition specifying the laws of nature that govern our world. Suppose that “H& L” implies that I will not go into my kitchen at time t_n+1for taking a glass of water. Actually, I decide not to go in my kitchen at time t_n+1. What if I did it? According to D. Lewis, while I could have not done it (on pain of denying that I could have broken a law), I could have been able to do something such that, if I have done it, a law would have been broken.⁸ This means that although a subject can not actually falsify a law of nature breaking its deterministic causal power, if the initial conditions of the world would have been different (remaining the laws of nature identical), at the time when in the actual world the subject performs action A, in the possible world with the different initial conditions she would have performed action B. Hypotetical freedom is the attribution to the subject of the potential ability to do something such that, if she did it, a law of nature would be broken. That is to say, hypotetical freedom expresses the idea that, given a plurality of worlds slightly different for some intrinsic conditions but identical as to the environment a determined subject lives in, the subject performs A in a world, B in another, C in a further one, and so on. This capture the human tendency to hold that a subject is able to do otherwise than what she actually does.

On the contrary suppose that at time t_n+1 I stand up and I move to my kitchen for taking my glass of water. In this case I carry out my resolution against the implication by “H & L”. Consequently, a law of nature would be broken (naturally, if the state of the world expressed by H has occurred as H states, H could not be falsified by event occuring after the occurrence of the state of the world expressed by H). Categorical freedom is the attribution to the subject of the actual ability to do something such that, if she did it, a law of nature is broken. That is to say, categorical freedom expresses the idea that, given a plurality of worlds slightly different for some intrinsic conditions but identical as to the environment a determined subject lives in, the subject performs A or B or C and so on in a world, A or B or C and so on in another, A or B or C and so on in a further one, and so on. This too capture the human tendency to hold that a subject is able to do otherwise than what she actually does.

Now, when Alciphron deduces I. (4) from the proposition that All in Nature is necessary determinedhe intends that, given that determinism is true, hypotetical freedom is irrelevant to human freedom. Indeed, since the kind of freedom usually attributed to human subject is categorical freedom, saying that a certain human subject could have done otherwise than what she actually does if she has fallen under a different causal chain, does not change the fact that this human subject is determined to do exactly what she does in the actual world and what she would do in any possible other. Consequently, necessary determined can be defined in the following manner:

Necessary determined =_def given a state of the actual world H occuring at time t_n, and a state of the world H’ occuring at time t_n in a possible world, and a further state of the world H’’ occuring at time t_n in a further possible world, (and so on), and the laws of nature L; X is necessary determined in the actual world iff “H & L” implies X at time t_n+1 and X actually occurs at time t_n+1, and X’ is implied by “H’ & L” at time t_n+1 in a possible world and X’’ is implied by “H’’ & L” at time t_n+1 in a further possible world, (and so on).

In my opinion Alciphron’s argument is a real challenge for Berkeley’s position. Because of the endorsement of the entire set of principle and premises of the argument, the only chance Berkeley has to defend Christianity from the freethinker’s attack is to reject I. (4). Furthermore, he doesneed to reply the freethinker’s attack: I. (4) appears to be a rebuttal of free will and the ascription of freedom to mankind is a foundational notion of the religion orientated conception of human existence Berkeley’s philosophy clearly aims at defending. Suppose, indeed, that I. (4) is true. From the assumption of a fourth premise follows another claim:

Premise₄: if a creature has not the possibility to act in a different way from how she actually acts, that creature is not free to choose to act in a way or another.⁹

Thus:

I. h) From Premise₄ it follows that if creatures do not have the possibility to act in a different way from how they actually act, then creatures are not free to choose to act in conformity with the content of their religious beliefs.

I. i) Suppose I. (4) to be true, it follows that men are not free to choose to act in conformity with the content of their religious beliefs.

I. j) But if men are not free to choose to act in conformity with the content of their religious beliefs, the traditional religious attitude to life is senseless.

By reason of this, Berkeley is forced at denying that I. (4) is true: his treatment of free will is mostly governed by the intention to defeat arguments such as Alciphron’s. Consequently, his doctrine of freedom depends on Alciphron’s understanding of the notion.

A very brief survey of the freethinker’s view: he provides three arguments against freedom, mainly drawn from freethinking literature.¹⁰ The first argument infers from the universal applicability of the laws of motion that all wordly occurrences are mechanically determined. Being mental events caused by wordly occurrences, mental events as volitions result determined too. The second argument infers from the evidence that volitions follow from moral judgements, that the mind is determined to will what it wills from causes external to the sphere of volition. The third argument infers from God’s omniscience that every event is a priori determined (if such and such cases will surely happen, then there are no contigent events).

It is evident that these arguments can not demonstrate I. (4). They simply conclude the trivial proposition that any determinate being subsists attending some others determinate beings, without asserting any clause about the nature of determination. Thus, in order to achieve a demonstration for the claim that no creature has the possibility to act in a different way from how she actually acts, Alciphron needs a principle stipulating the identity of determination and necessary determination. The following principle does the work:

Principle of Necessary Determination (PND): every determination is necessary determined.

That is, according to the definition of necessary determined, every actual determination follows from a given state of the world and the laws of nature, and every potential determination follows from states of the world of alternative possible worlds and the laws of nature.

In logical terms:

PND₁: ∀ x [P (x) ⊃ □ P (x)]

PND₂: ∀ x, y (x R y) {{P (x) ∧ [P (x) ⊃ Q (y)] } ⊃ □ {P (x) ∧ [P (x) ⊃ Q (y)] }}

The arguments then run as follows. Suppose will stands for y and the function Q for a modification of the will, i. e. a volition. Since volitions are always causally related to states of the world, let x stands for a certain state of the worldand R for a relationship of cause and effect. Alciphron argues for three different possibilities:

α) the function P expresses a mechanical property inhering in x;

β) the function P expresses a judgement whose content concerns x;

γ) the function P expresses a knowledge by God whose content concerns x.

In all these cases PND₂ asserts that since the will and a certain state of the world are related, if a given modification of that state of world implying a determinate modification of the will (or a given modification of human or divine mind whose content concerns that state of the world) occurs, volition results necessarily determined. Alciphron’s task consists therefore in providing reasons to relate volitions to modifications of determinate states of the world (or to modifications of human or divine mind whose content concerns determinate states of the world) in order to prove that volitions don’t stem from free will, being these necessary effects of external causes.

Berkeley’s strategy is to offer counterexamples to the adequacy of PND₂ to capture the relationship between volitions and modifications of states of the world (or modifications of human or divine mind whose content concerns determinate states of the world). As to the first argument, Berkeley refers to the ontological dualism of extended and unextended things in order to dismantle the identification of P (x) with some mechanical class of properties. Since objects belonging to the first group are completely different from those belonging to the second, there is no way to establish a direct relationship among them (Berkeley’s example is in my opinion quite unhappy: triangles and sounds). Now, any mechanical class of properties falls under the group of extended things, while volitions under unextended. Consequently, if P (x) ⊃ Q (y), under the premise that “x R y”, P (x) can not stand for any kind of mechanical class of properties.

As to the second argument, Berkeley denies that judgements and volitions stay in a relation of implication, since judgements and volitions are one and the same operation of mind; therefore, it follows that phenomena related to the relationship between judgments and volitions falls outside the extension of PND₂.

As to the third argument, Berkeley argues for the claim that while the conditional expressions recurring in the principle are formally true (if a knowledge by God whose content concerns a certain state of affairs occurs, then all minds experiencing that state of affairs will necessary enjoy those volitions God have foreseen), the premise “x Ry” fails to specify a causal condition, because certainty of knowledge is an epistemological requirement while necessity of occurrence is an ontological commitment.

At this point the notion of freedom Berkeley’s spokesmen and their adversaries deal with seems sufficiently clear: freethinkers and religion apologists focus the entire discussion on the determination of volition. While the first group appears confident in being able to prove that any volition is necessary determined, the latter rejects all arguments in support of the claim without denying that being free is being free to will. That is to say, Berkeley’s treatment of objections to the notion of freedom from freethinking literature highlights how matter at issue was informed by understanding freedom exclusively in terms of volition. In this regard, from a metaphysical point of view the agent’s will is the only principle by which accounting for agency. That’s why Euphranor judges free a man acting in conformity with his will. Since he supposes the will to be free, if an action flows directly from volition, i. e. if no external constraint prevents the agent from doing what she will, the action should be said free. On the contrary, since Alciphron supposes the will to be necessary determined, then the action flowing from volition couldn’t be actually free. Consequently, they both agree in defining freedom as correspondance of will and agency:

Freedom₁=_def “A man is said to be free, so far forth as he can do what he will” (Alc. VII. 22, p. 263).

Freedom₂=_def “Man acting according to his will, is to be accounted free” (Ibidem).

Although both definitions appear really traditional, the debate between Berkeley’s spokesmen and their adversaries is very instructive as to the identification of human freedom with freedom of choice. Indeed, their common assumption that determinismwould dismantle any attribution of freedom to mankind shows a substantive reason in support of claim as (C₁). Given the understanding of determinism in terms of Necessary determined=_def and PND₂, human agency should appear the outcome of the states of the world and the laws of nature. Why does this conclusion challenge the idea that human being are free? After all it could be said that an agent acts freely whenever he acts with no external constraints, even if her actions are necessary determined.

What makes this proposal unsatisfactory is the inconsistency between the notion of acting with no external constraints and acting by being necessary determinedwhen applied to moral agents as human subjects are supposed to be. Indeed, if an agent acts by being necessary determined then her actions are the outcome of states of the world and the laws of nature. But if this is the case, the agent does not act with no external constraints. On the contrary, if the agent acts with no external constraints, then her actions should appear determined by some ability to perform an action with which the agent is essentially equipped. That is, the agent should acts basically in reason of her internal principles. Therefore, conclusion follows: if determinism is true there is no possibility to reject the idea that every human subject is necessary determined to act by some kind of external constraints.

Now, this conclusion makes the pair with the rejection of the idea that freedom could be attributed to human being, because human freedom is commonly meant to be the possibility to perform an action in reason of the agent ability to determine herself in acting.¹¹ The reduction of freedom to freedom of choice intervenes here: the will is the source of the agent’s ability to determine herself to act, since an agent determining herself is an agent willing herself to behave in a certain manner. That is to say, the subject is not necessary determined by external constraints in acting because she has a special power to decide how to act and to actualize this resolution by self-determination: a subject free in doing what she does is an agent doing what she will and willing what she likes.

A promising way to deploy this intuition is to say that to be free a human subject should possess immanent causal powers. According to R. Chisholm causal powers are divided into two main categories:¹²

… when one event or state of affairs (or set of events or states of affairs) causes some other event or state of affairs, then we have an instance of transeunt causation. … when an agent, as distinguished from an event, causes an event or states of affairs, then we have an instance of immanent causation.

This distinction makes appeal to the phenomenal evidence that, although states of the world succeed the one to the other appearing causally chained in a train of events, human subjects seem to be endowed with a special, basic and irreducible power to start new chained trains of events. Given that the world we live in appears a chain of transeunt causes, agency theorists (theorists attributing to human subject immanent causal powers) need to answer how immanent causation can work within a transeunt causation realm. The usual idea is to express immanent causation in terms of control of transeunt causation processes.

Quoting Chisholm again:

The nature of what is intended by the expression immanent causation may be illustrated by this sentence from Aristotle’s Physics: “thus, a staff moves a stone, and is moved by a hand, which is moved by a man” (VII. 5. 256a, 6-8)… We may say that the hand was moved by the man, but we may also say that the motion of the hand was caused by the motion of certain muscles; and we may say that the motion of the muscles was caused by certain events that took place within the brain. But some event, and presumably one of those that took place within the brain, was caused by the agent and not by other events.

Berkeley’s reference to self-determination as something that is exercised by the will provides a theoretical model for the explanation of human immanent causal powers. This model characterizes human agency as the production of “sighted actions”, namely, intentional (or orientated-by-the-will) actions the agent performes with some goals in mind.¹³ That is to say: states of the world and the laws of nature determine events to occur. The agent experiences those events acquiring some packs of informations about her actual environment. Then the agent responds to the state of the world by deciding which purpose to pursue with her behaviour, that is, being directed towards the achievement of something she will. Consequently, human subjects exercise control over their action by their will, being in this way responsible for what they do.

In the light of this agreement between Berkeley’s spokesmen and their adversaries it seems reasonable to expect that the relationship between freedom and the will gives Alciphron further refined reasons in support to his argument. Indeed, he requires further conditions for saying a man to be free: if the will is the principle accounting for agency, it seems plain to claim that a man is free if and only if he is free to will what he will. According to the freethinker this means that the will is free if and only if the will is completely free to self-determine a volition among (at least) a pair of alternative possibilities. He says:

In freedom there should be an indifference to either side of the question, a power to act or not to act, without prescription or control: and without this indifference and this power, it is evident the will cannot be free (Alc., VII. 20, p. 260).

This requirement of indifference in choice means that, given Freedom₁=_def and Freedom₂=_def, according to the freethinker freedom can be attributed to human subject if and only if the following condition obtains:

Freedom₃=_def a person is free to will in willing what he does iff he could have will otherwise.

It seems, then, that Alciphron assumes as the default option concerning free will a notion that appears to be the general form of what H. G. Frankfurt has called the Principle of Alternate Possibilities (PAP):¹⁴

PAP=_def A person is morally responsible for what he has done only if he could have done otherwise.

Indeed, given that Alciphron and Euphranor agree on the claim that the attribution of freedom to mankind implies that human subjects are agents provided with an immanent causal power they exercise by their will, according to both of them an agent is morally responsible for her actions if and only if the agent freely will the actions she performs. That is, under these common assumptions the notion of moral responsibility requires the notion of free will.

Because of this, a stronger argument appears to be at Alciphron’s disposal. If the notion of freedom excludes any kind of determination in volition, he can express a very malicious doubt (Alc., VII. 22, p. 263). Suppose a man to be free because of his acting in conformity with his will. Would he be really free to will what he wills? That is, would his will really be able to self-determine the given volition choosing among (at least) a pair of alternative possibilities? If the answers to these questions are negative, as Alciphron evidently holds, the supposition that the will is free to self-determine volitions would not suffice in order to defend the attribution of freedom to mankind. Indeed, even if the will self-determines volitions, this self-determination could be necessary determined. All this is captured in my opinion by PND₁: Alciphron’s doubt is then an evidence for its assumption.

Naturally, PND₁ is a basic metaphysical principle, that is, a first evidence whose analisys prima faciecan’t refer to more basic principles. Consequently, it results extremely difficult to provide a conceptual argument against PND₁ without assuming some experiential intuitions contrary to its content. In my opinion this is the main reason why Berkeley’s positive argument for free will simply appeals to the inward experience of volition, rejecting any intention to argue conceptually for it:

If I should suppose things spiritual to be corporeal, or refine things actual and real into general abstracted notions, or by metaphysical skill split things simple and individual into manifold parts, I do not know what may follow: but if I take things as they are, and ask any plain untutored man, whether he acts or is free in this or that particular action, he readily assents, and I as readily believe him from what I find within (Alc., VII. 21, p. 262).

Now, in spite of Berkeley’s confidence in referring to commonsense intuitions, his claim seems to me really puzzling. If moral agents are thought to be free because they have immanent causal powers relying on their will, then Alciphron’s requirements for attributing freedom to mankind should be answered. As to this, Berkeley’s claim seems to me senseless: my “plain” experience, indeed, makes me believe I’m not usually free in willing what I will. That is to say, a vague reference to ordinary experience of the production of volitions by the will does not seem to provide a positive case in support to the claim that human beings are free.

2.

In order to explain why, I begin with mentioning some facts occuring to me when I was younger. After the fulfillment of my secondary studies, at the age of 18 I begin to smoke cigarettes. I’ve happily been a heavy smoker for a period of three years. I remember quite well that I didn’t care about consequences nor I had particular problems with nicotine. When I was smoking, I simply felt a strong pleasure. Nonetheless, things began to change since I had realized that I was completely cigarette addict. I came at this awareness because much of the cigarettes I smoked didn’t give me pleasure anymore, even if some continued to satisfy me (the first in the morning, after lunch and dinner, the last before sleeping). Consequently, I tried to set a limit to my smoking, but I couldn’t. Although I really wanted to smoke just those cigarette that I desired, I can’t avoid smoking my usual quantities. Therefore, I took a resolution and I quit smoking.

Now, I perfectly remember that during working hours I often had cigarette breaks. In the period of my empty efforts to limit my cigarettes, I didn’t like to smoke in those moments, but I smoked the same. Before lighting my cigarette I usually thought of all the unpleasant feelings I would have felt, but I finally decided to light it. It is worth noting that I didn’t do it suddenly. It was as if a part of me didn’t want to smoke, while another did. Since my act was the result of my choice, I’m inclined to think that my will produced a volition to smoke. Nonetheless, I felt this volition as something which does not belong to my volitional life, since it always went with another contrary volition (whose content was I do not want to smoke) that seems to me to express more properly what I did will. In such a case, I was sure that the force of these volitions was exactly the same: the fight was solved just by my habit to smoke. For this reason I can’t express myself differently than saying that I would have smoke without willing.

In my opinion such situations are very common. Each of us, I claim, has experienced strong hesitations due to the will of performing two different and opposite actions, the first being the purpose of a compulsory volition, the second the purpose of a contrary volition which appears to express our actual will. When the first of them is performed, the feeling enjoyed by the agent is that she will without willing (or that she will without being free in willing). The main reason for this odd feeling is, I think, that in similar cases the moral conflict is mostly lead to its solution by custom or by chance, i. e. the performance of the action is not determinated by the will, being determined by something external the opposition of volitions. Consider the following situation: my will produces two opposite volitions, V₁ and V₂. I should making a choice among them. Nonetheless, neither of them prevails since they are equally compelling. Consequently I can not choose V₁ nor V₂. Suppose I finally determine myself to perform the action corresponding to the content of V₁. Since V₁ and V₂ have the same volitional force, the choice has occurred by means of motives other than the will. Be these motives custom, chance or whateverelse: since I would both V₁ and V₂, performing the action corresponding to the content of V₁ I must be said to have would the action I perform. Nonetheless, the action corresponding to the content of V₁ hasn’t been freely determined by the will: willing the action has been determined by the motive inclining me to perform it; which is the same with saying that the will is not always freely self-determined in willing.¹⁵

Phenomenal evidences of this kind raise a fundamental difficulty for a defense of human freedom moving from assumptions as (C₁) and the recourse to experiences of choices among different possibilities as positive cases in support to the freedom of the will to determine itself. Indeed, consider the following argument:

II. a) Given determinism, necessary determined=_def and PND₂, it follows “Freedom₁=_def & Freedom₂=_def”; that is, (C₁) is the right understanding of human freedom;

II. b) Given “Freedom₁=_def & Freedom₂=_def”, it follows Freedom₃=_def; that is, since (C₁) is the right understanding of human freedom, then attribution of freedom to human beings implies that human freedom should be free to will;

II. c) Phenomenal evidences show that, at least for cases such as hesitations, human beings are not free to will.

According to the argument II. (a)-(c), phenomenal evidences provide counterexamples to the claim that human beings are free to will. Naturally, it could be said that (C₁) and Freedom₃=_def apply to those cases of human agency which are relevant to the attribution of freedom to mankind, wheras other cases of human agency, e. g. hesitations, could consistently appear to be not free. Nonetheless, the argument could be refined in a way that it results almost unanswerable.

Consider the reasons why Berkeley and the agency theorists make appeal to the notion of (free) will: if determinism is the whole story concerning the succession of states of the world, then the subject’s actions are out of her control. But the attribution of freedom to a human subject implies that the human subject has some kind of control (or authorship) over her actions. (Free) will, namely, freedom of choice, provides phenomenal evidence that human subjects exercise control over their actions. Therefore, determinism is not the whole story concerning the succession of states of the world.

In reason of this it appears evident that Berkeley and agency theorists resort to (free) will as a mean to exercise immanent causal power (to lead agent causation) on transeunt caused successions of states of the world (on event causation processes). Common examples of this ability concerns ordinary situations such as performing different kind of particular actions,¹⁶ probabilistically deciding (by weighted motives between a pair of possibilities) how to spend time;¹⁷ resolving moral conflict among opposing interests.¹⁸

What if in all these situations the will would result determined by something other than itself? I have a strong inclination to hold that this matter stands as follows: the agency theorist’s examples relate to situations where the subject need to choose among two or more (conflicting or non conflicting) possibilities. A gross amount of classical researches by psychologists and sociologists attest that choice of this kind are mostly governed by habits, custom and compliance to the social role the subject play. Consequently, the phenomenal evidences referred to in II. (c) can be justifiably thought a subset of the phenomenal evidences involved in every exercise of (free) will. The argument II. (a)-(c) is then to be reworked as follows:

II. d) Given determinism, necessary determined=_def and PND₂, it follows “Freedom₁=_def & Freedom₂=_def”; that is, (C₁) is the right understanding of human freedom;

II. e) Given “Freedom₁=_def & Freedom₂=_def”, it follows Freedom₃=_def; that is, since (C₁) is the right understanding of human freedom, then human beings should be free to will;

II. f) Agency theorists appeal to exercises of (free) will in choosing among two or more (conflicting or non conflicting) possibilities as phenomenal evidences in support to the actual soundness of Freedom₃=_def in capturing the notion of human freedom;

II. g) Phenomenal evidences show that human beings are not determined by (free) will in choosing among two or more (conflicting or non conflicting) possibilities;

II. h) Human beings are not free to will.

Now, given that determinism, necessary determined=_def and PND₂ does not seem prima facienegotiable and that, on these assumptions, the attribution of freedom to mankind requires (C₁) to be valid, if a theorist aims at defending the attribution of freedom to mankind, then arguments as II. (d) — (h) should be dismantled.

In my opinion the most promising way to accomplish the task is to give an account of the will in terms of the constitutive theory of self-knowledge:¹⁹ this approach allows to model volitional events relying on the distinction among two basic sets of volitions; which suffices to reject the conclusion II. (h).

Self-knowledge is the first person knowledge of one’s mental states as beliefs, desires, and sensations. Constitutivism concerning self knowledge is the claim that having knowledge of one’s first order mental states is at least a necessary condition for having those mental states (CSK). The main reasons in support of CSK are due to transparency (or immediacy) and authority. Suppose a subject has self knowledge that she believes that P. What does this mean? The natural answer is that she believes that P. Transparency (or immediacy) captures the idea that self knowledge of one’s first order mental states necessary goes with (is one and the same with) the occurence of those first order mental states. Suppose further that someone asks that subject how can I say that you believe that P?. The subject should answer you can say that I believe that P since I testify to you that I do believe that P. Authority grasps the idea that, given a mature and healthy subject, self knowledge of one’s first order mental states is justified by the first person enjoyment of those mental states.

The main problems for the assumptions of CSK relate to two classes of phenomenal evidences that provide counterexamples to the adequacy of CSK. The first class consists in unconscious mental states; the second in the purposive behaviours of animals lacking self knowledge. Both of these classes of phenomenal evidences address to CSK the same charge: at least basic mental states as beliefs, desires, and sensations possibly occur independently on having knowledge of them.

The constitutive account of self knowledge rejects the objection from unconscious mental state and purposive behaviour introducing the distinction between beliefs as commitmentsand beliefs as dispositions. The first are judgement-sensitive: they express propositional knowledge, which commits the subject of belief to hold responsibly proposition as I believe that P. The second are non-judgement-sensitive: they express non propositional knowledge, which does not commit the subject of belief to hold responsibly proposition as I believe that P, even if they dispose her to act in conformity to such a belief. Unconscious mental states and purposive behaviour are related to the field of application of the notion of beliefs as dispositions, while CSK applies to beliefs as commitments.

Let’s now turn to the will and volitions. My proposal is to deal with willing as a particular mental state falling under the extension of CSK. Prima facieit seems extremely reasonable, since both transparency and authority appear to be properties of volitional mental states. Indeed, having knowledge of one’s will (or volitions so and so) necessary goes with (is one and the same with) the enjoyment of one’s will (or volitions so and so). And again, self knowledge of one’s will (or volitions so and so) is justified by the first person enjoyment of one’s will (or volitions so and so). Consequently, it can be said having knowledge of one’s will (or volitions so and so) is at least a necessary condition for having one’s will (or volitions so and so).

In order to answer the counterexamples that hesitations, conflicting volitions and non conflicting volitions provide to the claim that human beings are free to will, I parallel the distinction between beliefs as commitments and beliefs as dispositions in relation to volitions. I say that volitions as commitments are judgement-sensitive volitions expressing propositional knowledge such that, if a subject enjoys a given volition of this kind, then she holds responsibly proposition as I will so and so. And again, volitions as dispositions are non-judgement-sensitive volitions expressing non propositional knowledge such that, if a subject enjoys a given volition of this kind, then she does not hold responsibly proposition as I will so and so, even if she results to be disposed to act in conformity to such a volition.

It seems to me that this approach defeats the argument II. (d) — (h). This is how I see the story: while the will produces consciously volitions as commitments, the will arranges unconsciously volitions as dispositions by simply confirming the behavioural persistance of subjective habits, aptitudes and inclinations. Now, the will is free in willing the first set of volitions; while is not free in willing the second (the will just deliberates that if such and such is the situation, then such and such should be the subject’s behaviour). As regards to this distinction, I claim that Freedom₃=_def does not apply to all kind of agency human subjects ordinary experience; being its extension limited to volitions as commitments.

In this light hesitations, conflicting volitions and non conflicting volitions are fights between a volition as commitment (which appears to the subject soundly expressive of her will) and a volition as disposition (which appears to the subject unsoundly expressive or basically inexpressive of her will). As a consequence, the assumption of II. (f) in the argument is faulty: agency theorists fail to identify the right exercises of free will in support to Freedom₃=_def. Then, the soundness of II. (g) is not a case against Freedom₃=_def.

With all these considerations in mind the problem of freedom can now be newly formulated. The appeal to common sense experiences of acting reveals to be disappointing since the most part of human agency probably concerns just the fight between volitions as commitments and volitions as dispositions, which does not appear the field within which human beings exercise their freedom. Nonetheless, human beings possibly have an experience of their freedom in actualizing volitions as commitments. The attribution of freedom to mankind, then, is to be defended by the right identification of cases wherein the will produces volitions as commitments.

In my opinion Berkeley’s theory of agency provides a preliminary point of departure towards a palatable understanding of the topic. Indeed, his account makes justice both to the claim that human beings commonly act by custom and to the claim that human beings can be free in acting. Naturally, I’m not saying that Berkeley offers us a consistent and complete theory of freedom: I simply hold that, although he does not deal with freedom and agency by the adequate conceptual distinctions and the sound systematic architecture which should be required (for example, Berkeley does not develop a persuasive account of the different kind of volitions), the intuitions on which his embryonic version of TPA relies deserve to be worked out into a consistent and complete theory.

3.

In order to set forth Berkeley’s embryonic approach to TPA, I will begin with focusing on a passage (Alc., I. 11, p. 40) wherein Berkeley briefly sketches out a theory of agency. In my opinion the text is very relevant because Berkeley does not approach the topic from a systematic point of view: since he treats a practical case of agency, the passage shows Berkeley’s conception at work.

The matter at issue concerns whether a technical method in knowledge is useful or useless. Alciphron and Lysicles defend the claim that in knowledge “method, exactness and industry are a disavantage”. Refuting academical methods of inquiry, based on sistematical reading, formal arguing and disciplined meditation on antinomic views, freethinkers believe freeconversation or speaking to be the most suitable mean to achieve the truth. In support to this, Alciphron refers to the difference between the Dutch and the Italian manner in painting. According to his opinion, while the former consists in applying established techniques to picture exactly the objects represented, the latter appears to be the outcome of a free inspiration relying on emotions, intuitions, and enthusiasm. That is, he holds that the example highlights how proceeding freely warrants better results than proceeding methodically. Consequently, he invites Euphranor to apply this consideration to the claim he is defending. Berkeley’s spokesman replies as follows:

… did those great Italian masters begin and proceed in their art without any choice of methods or subject, and always draw with the same ease and freedom? Or did they observe some method, beginning with simple elementary parts, … which they drew with great pains and care, … in order to draw them correctly, and so proceeding with patience and industry, till after a considerable lenght of time they arrived at the free masterly manner you speak of (Alc., I. 11, p. 40)?

Opposing the attitude of freethinkers to hold that free manneris the ability to produce actions floating arbitrarily by association of ideas or feelings,²⁰ Euphranor suggests that the exercise of human agency requires to acquire behavioral automatisms.

A line of reasoning justifies the soundness of this suggestion: human beings have natural inclinations (as sensations, passions, fears, appetites, aversions and desires) which direct them to pursue different ends. These inclinations oppose and fight the one to the other, being the one strongest at a certain moment and weaker at another, the other weaker at a certain moment and strongest at another. Consequently, human beings are naturally driven to behave in a way or another in reason of the random result of the fight among their natural inclinations, insofar as they follow these.²¹ But, human beings are something more than natural animals: they are moral agents, that is, they have rationality and judgement in order to direct their behaviours. Nonetheless, their rationality and judgement have the unavoidable tendency to fall into a theoretical and practical failure. As to the theoretical side: human reason is constantly defeated by faith, prejudice and unjustified beliefs.²² As to the practical side: human reason appears unable to subdue completely the fight among natural inclinations to the actualization of the decrees of the will.²³ In consequence of this double failure, Berkeley rejects a conception of deliberation as an exclusively rational balance among moral reasons:

Whatever may be the effect of pure theory upon certain select spirits, of a peculiar make, or in some other part of the world; I do verily think that in this country of ours, reason, religion, law, are all together little enough to subdue the outward to the inner man (Alc., III. 12, p. 107).

Now, phenomenal evidences as hesitations, conflicting volitions and non conflicting volitions seems to me reasons in support to Berkeley’s account of human moral life. In light of the distinction among the two basic sets of volitions I have introduced in the previous section, I claim that, while the subject ordinary performs actions being driven by persistently prevailing volitions as dispositions (in Berkeley’s terminology: natural inclinations), whenever her actions result relevant for the self-ascription of control or authorship over these, the subject performs actions by means of an effort to win the persistently prevailing volitions as dispositions by her volitions as commitments (in Berkeley’s terminology: rationality and judgement).

If this is the case, this picture of human agency conflicts with the assumptions (c), (d), and (e) by which the contemporary debate on freedom characterizes the nature of volitions and actions. Indeed:

Refutation of (c): the volition forming process is not to be referred to the subject only, since volitions as commitments are framed by the interaction of reasons as moral judgements, unjustified beliefs as prejudices and natural inclinations as volitions as dispositions. Moral judgements are in the subject’s complete control, whereas prejudices and volitions as dispositions are not. The same holds for the action forming process, since either the performance of an action ordinary depends on the exercise of volitions as dispositions, which does not result in the subject’s complete control, or the performance of an action relevant for the self-ascription of control (or authorship) over this, depends on the effort to subdue volitions as dispositions to volitions as commitments, which is not always the case.

Refutation of (d): the volition forming process and the action production process are not linear processes, since in both of them occur motives different in kind (moral judgements, prejudices, volitions as dispositions). Consequently, these can not count in the processes by their absolute weight or force (for example, a strong judgement as you shall not use violence to your lover could win a contrary weaker judgement as you could use violence to your lover if she makes you get angry, even if it could not be able to subdue a natural inclinations to wrath and violence).

Refutation of (e): the volition forming process and the action forming process do not occur instantaneously since they are both the result of temporal interactions among reasons different in kind (for example, a subject naturally inclined to wrath and violence could possibly be convinced that she should not use violence to her lover, so that she could possibly fight among her natural inclinations during a fit of anger in order to subdue these).

Berkeley’s appeal to the acquisition of behavioural automatisms matters here. Given that human pratical life reveals to be subjected to the random swinging among different purposes, the agent is exposed to the risk of wandering through the achievements of eteronomical ends, being driven from the one to the other by the fight among motives different in kind. The only mean the subject has to exercise control (or authorship) over her actions seems then to conform her behaviour to a customary performance of sets of actions whose actualization is made possible by the acquisition of their execution by training.²⁴ That is, the subject need to model her behaviour by making her volitions as commitments to lead the processes by which persistently volitions as dispositions are established, reinforced and dismantled.

For example, consider the case Alciphron and Euphranor deal with in Alciphron I. 11. Suppose a painter intends to portray a beatiful woman. In order to accomplish this task she should be able to perform determined actions: handling paintbrushes, mixing primary colours, using degrees of colour to achieve perspectival effects, and so on. If this is the case, she should then be able to apply these tecniques in picturing hands, picturing faces, picturing legs, and so on. And again, if she can adequately picture different parts of the human body, she should be able to vary them in different postures. Each of these steps is customary acquired by performing several resembling instances of the same action, to such an extent that, whenever the action should be performed again, then the action is automatically performed. This acquisition is to be seen as the establishment, reinforcement and dismantling of volitions as dispositions (consisting in natural inclinations towards the exercution of a given performance of handling paintbrushes, of mixing primary colours, of using degrees of colour to achieve perspectival effects, of picturing hands, of picturing faces, of picturing legs, and so on) by volitions as commitments (consisting in judgements such as I will determinately handling paintbrushes so and so, I will determinately mixing primary colours so and so, I will determinately using degrees of colour to achieve perspectival effects so and so, I will picturing hands so and so, I will picturing faces so and so, I will picturing legs so and so, and so on). Finally, since the painter controls all these particular actions, she appears able to apply any of them in portraying the beatiful woman.

Summarizing, a berkeleian-like account of free agency consists in the conjunction of five different intuitions:

I_a) Human beings are free because they have a control or authorship over some of their actions;

I_b) The mean by which human beings exercise control or authorship over their relevant actions is their will;

I_c) The will is not able to exercise an absolute power to produce volitions and direct the subject’s agency, because, given that the practical life is a fight between moral reasons, prejudices and natural inclinations, volitions as dispositions ordinary seem strongest than volitions as commitments;

I_d) The will does exercise a relative power to actualize its volitions as commitments by making them to lead the customary acquisition of habits by training;

I_e) Human beings are free in their behaviour iff the exercise of their agency by their volitions as dispositions is the result of an intentional customary acquisition by training of behavioural automatisms.

Since from the assumptions of (I_a), (I_b), (I_c), (I_d) and (I_e) follows that (c), (d) and (e) should be rejected, a berkeleian like account of free agency appears to be irreducible to any of the views (1), (2), (3), and (4) the contemporary debate on freedom admits. That is to say, since Berkeley’s approach to TPA relies on a characterization of volition and action which is alternative to that assumed by (1), (2), (3), and (4), a berkeleian like account of free agency essentially differs from each of the contemporary views on freedom.

Indeed, consider libertarianism. Both the libertarian theorist and Berkeley are agency theorists. Nonetheless, the first holds, while the second denies, that the will exercises control over the subject’s relevant actions by having the power to choose among alternative possibilities and by having the power to actualize her choice. That is, the libertarian relies on an optimistic account of the causal powers of the will; Berkeley on a pessimistic one.

Let’s now move to hard-determinism. Both the hard determinist theorist and Berkeley claim that human agency is ordinary determined by external constraints on the subject. Nonetheless, the first holds, while the second denies, that this suffices to reject the attribution of freedom to mankind. That is, the hard determinist thinks of freedom in terms of autonomy in willing and acting; Berkeley in terms of control over the process leading to choice and execution of actions.

As to compatibilism. Both the compatibilist theorist and Berkeley hold that a deterministic approach to human agency can be made consistent with the attribution of freedom to mankind. Nonetheless, the first holds, while the second denies, that the ascription of authored actions to a given subject by other subjects is enough to attribute moral responsibility to her. That is, the compatibilist assumes that the attribution of freedom to human beings is justified by common social practices; Berkeley that the attribution of freedom to human beings requires they to be metaphysically responsible for their behaviour.

Finally, I come to skepticism. Both the skeptic theorist and Berkeley think that any theory being a version of (1), (2), and (3) should be rejected. Nonetheless, the first holds, while the second denies, that, given the rejection of (1), (2), and (3), no theory can adequately grasp the notion of human freedom. That is, the skeptic is ready to argue for the lack of soundness of any speculation on freedom; Berkeley for the soundness of his own one.

Consequently, a berkeleian like account for free agency appears a case for the assumption of (C₂). Now, relying on Berkeley’s account of moral life and his intuitions about customary agency by acquisition of behavioural automatism, I will sketch an explanation of acting as a theory of procedural agency. This theory understands the performance of actions by an agent in terms of automatical performance of customary acquired subactions. I think a provisional version of TPA can be captured by the following propositions:²⁵

Requirements for acquisition of behavorial automatisms

R₁) An agent acquires performing any determined action by custom;

R₂) An agent acquires how to perform any determined action as a determined instance of a determined type of action;

R₃) An agent is able to perform an action iff she has acquired to perform that action.

Definition

For any action a_i, subactions of ai are all instances a¹, b², . . ., k^k+a of A, B, . . ., K whose performance is required for performing a_i.

Principle of performance of actions

PPA) An agent performs an action ai performing groups of subactions of a_i.

Theory of procedural agency

TPA₁) An agent performs an action a_i iff she has acquired to perform the subactions of a_i.

TPA₂) If an agent performing the action a_i performs the subactions b₂, … ., k_k+a, then b₂ is a particular instance of b-type action, …, k_k+a of k-type action.

TPA₃) An agent performs an action a_i performing groups of customary acquired subactions, which are particular instances of determined type of actions.²⁶

In my opinion TPA is the most promising approach to freedom since, while the theory grants that human agency has a predominant deterministic feature, it rightly identifies the class of phenomena which makes justice to the idea of dealing with freedom in terms of the exercise of a free agency. That is, TPA finds phenomenal evidences in support to the claim that human beings are free in exercising free agency: insofar as subjects model their behaviour by educating themselves to perform customary behavioural automatisms, they freely commit themselves to act determinately in the way they like.

What matters in TPA is that it enables the theorist to account for the claim that human beings have immanent causal powers within a transeunt causation realm. Consider a subject experiencing a given state of the world. Normally, she answers environmental stimuli by her volitions as dispositions. That is to say, the subject reacts to the situation the actual state of the world presents her by her natural tendencies to perform the action so and so whenever the situation she experiences is so and so. Now, both the environmental stimuli and the subject’s reaction to these are a clear case of transeunt causation processes. Nonetheless, the subject’s reaction is a customary performance of acquired behavioural automatism. Immanent causation by the subject intervenes here: the acquisition of behavioural automatism is the establishment, reinforcement and dismantling of persistently prevailing volitions as dispositions. That is, human will models the subject’s reaction to a given state of the world by her volitions as commitments. Consequently, the subject’s reaction to a given state of the world does not originate from any transeunt causation process, being the result of the effort by the will to subdue volitions as dispositions to volitions as commitments. Conclusion follows: the more the subject conforms her volitions as dispositions to her volitions as commitments, the more the subject exercises control over her behaviour; namely, the subject exercises immanent causation powers.

Naturally, this does not mean that the preliminary version of TPA is a sufficient and complete account of freedom: it looks more like a starting point than a definitive theory. Indeed, in order to accomplish the task there is a great deal of problems that requires to be faced. In my final section I will briefly outline what seems to me in need to be settled.

4.

Problems for TPA consist in giving a sustained theory of the will addressing mainly metaphysical concerns. This is the way I see this matter.

TPA should be able to expand its fundamental intuition on the way the will exercises control over the subject’s agency. That is, it is necessary to provide a detailed account of the manner by which volitions as commitments and volitions as dispositions originate. Furthermore, the account should show how it is possible that volitions as commitments lead the establishment, the reinforcement and the dismantling of volitions as dispositions.

Once this account is given, in order to overcome antinomic reasoning, TPA should reject (1), (2), (3), and (4) with ad hoc arguments relying on its main propositions. The same holds for any version of the regressum rationiargument or the luck principleargument.

The first argument runs as follows:²⁷

IV. a) When a subject either will or acts, she either determinately will or determinately acts;

IV. b) Therefore, when a subject will or acts, she either will or acts given a determined situation;

IV. c) That is, given the determined situation, the subject will or acts reacting to the determined situation;

IV. d) Now, the subject will or acts being either free or not in exercising control over her volitions or actions;

IV. e) The possibility that the subject is not free is not the case for the vindication of the claim that human being are free in exercising free agency;

IV. f) Consequently, suppose she is free because she reacts exercising control over her volitions or actions;

IV. g) Then, she has some motives in her control for determining herself to behave in a determinate manner;

IV. h) A motive determining a subject to behave in a determinate manner is in her control, iff the subject is committed to behave in a determinate manner;

IV. i) A subiect is committed to behave in a determinate manner for some reasons;

IV. j) Consequently, if the subiect will or acts exercising control over her volitions or actions, then she has some reasons committing herself to behave in a determinate manner;

IV. k) Now, if the subject has some reasons committing herself to behave in a determinate manner, either she holds justifiably those reasons or not;

IV. l) If she does not hold justifiably those reasons, then the subject has not any actual reason to hold those reasons;

IV. m) Consequently, the subject will or acts without having any actual control over her volitions or actions;

IV. n) On the contrary suppose she has some second order reasons to hold those first order reasons;

IV. o) Then she should have some other third order reasons in order to justified those second order reasons, and so on;

IV. p) Finally, in having some higher order reasons justifying all her lower order reasons, any subject falls in an infinite regressum rationi;

IV. r) Now, given that human subjects are finite beings, they can not perform an infinite regressum rationiin order to justify those reasons which commit them to behave in a determinate manner;

IV. s) Therefore, the subject can not will or act exercising control over her volitions or actions.

The second argument is someway related to the first argument. Suppose that a theorist wishes to defeat the regressum rationi argument by denying that (IV. a), (IV. b), and (IV. c) are sound. Therefore, it should follow that (IV. t) the subject will or acts being undeterminately free to will or act otherwise than what she actually does. Let’s assume the following principle:²⁸

Luck Principle=_def If an action is undetermined at time t, then its happening rather than not happening at t would be a matter of chanceor luck, and so it could not be a freeand responsible action.

Then the argument follows:

IV. t) the subject will or acts being undeterminately free to will or act otherwise than what she actually does;

IV. u) The Luck Principle is the case;

IV. w) The subject is not free in either willing or acting.

If all these purposes be achieved, TPA would result a consistent and complete account of freedom. I think much work is required to this end. Nonetheless, TPA appears enabling theorists to achieve the goal of vindicating the idea that human beings are metaphysically free in a very promising way.

Bibliography

5.1. Primary Sources

• L. Jaffro, G. Brykman, C. Schwartz (2009), eds., Berkeley’sAlciphron. English Text and Essays in Interpretation, Hildesheim-Zürich-New York: Georg Olms Verlag.

5.2. Secondary Sources

• Berman, D., (2009), A Manual of Experimental Philosophy, Dublin: Jeremy Pepyat Books.
• Bertini, D., (2005), “L’immaterialismo berkeleiano come teoria religiosa dell’esperienza” in D. Bertini, ed., George Berkeley. Alcifrone, ossia il filosofo minuzioso, Milano: R. C. S. Libri, pp. 7-100.
• Bertini, D., (2010), “La critica berkeleiana dell’autonomia morale” in L. Jaffro, G. Brykman, C. Schwartz (eds.), Berkeley’sAlciphron. English Text and Essays in Interpretation, Hildesheim-Zürich-New York: Georg Olms Verlag, pp. 309-318.
• Brykman, G. (2009), “La liberté humaine dans le dialogue VII de l’Alciphron”, in L. Jaffro, G. Brykman, C. Schwartz (eds.), Berkeley’sAlciphron. English Text and Essays in Interpretation, Hildesheim-Zürich-New York: Georg Olms Verlag, pp. 331-340.
• Chisholm, R. M. (1964), “Human Freedom and the Self”, in G. Watson (ed.), Free Will, Oxford-New York: Oxford University Press, pp. 26-37.
• Clarke, R. (1993), “Toward a Credible Agent-Causal Account of Free Will”, Nous, 27, pp. 191-203.
• Coliva, A. (2009), “Self Knowledge and commitments”, Synthese, pp. 365-375.
• Fischer, J. M. (1994), The Metaphysics of Free Will: An Essay on Control, Cambridge, Mass.: Blackwell Press.
• Frankfurt, H. G. (1969), “Alternate Possibilities and Moral Responsability”, Journal of Philosophy, 66/23, pp. 829-839.
• Harris, J. (2009), “Berkeley on the Inward Evidence of Freedom”, in L. Jaffro, G. Brykman, C. Schwartz (eds.), Berkeley’sAlciphron. English Text and Essays in Interpretation, Hildesheim-Zürich-New York: Georg Olms Verlag, pp. 331-340.
• Jaffro, L. (2003), “La paideiaselon Berkeley et Shaftesbury: un conflit dans la conception thérapeutique de la philosophie”, Le Temps Philosophique, 9, pp. 147-173.
• Lewis, D. (1981), “Are We Free to Break the Laws? ”, Theoria 47, pp. 112-121.
• Kane, R. (1999), “Responsibility, Luck, and Chance: Reflections on Free Will and Indeterminism”, Journal of Philosophy, 96/5, pp. 217-240.
• Muehlmann, R. G. (1992), Berkeley’s Ontology, Indianapolis-Cambridge: Hackett Publishing Company.
• O’Connor, T. (1995), “Agent Causation”, in T. O’Connor (ed.), Agents, Causes, and Events: Essays on Indeterminism and Free Will, New York-Oxford: Oxford University Press, pp. 173-200.
• Olscamp, P. J. (1970), The Moral Philosophy of George Berkeley, The Hague: Martinus Nijhoff.
• Smart, J. J. C. (1961), “Free Will, Praise and Blame”, Mind, 70, pp. 291-306.
• Strawson, G. (1994), “The Impossibility of Moral Responsibility”, Philosophical Studies, pp. 5-24.
• Strawson, P. (1962), “Freedom and Resentment”, The Proceedings of the British Academy, 48, pp. 1-25.
• Van Inwagen, P. (1983), An Essay on Free Will, Oxford: Oxford University Press.