Free will, indeterminacy, and self-determination

Table Of Contents

Introduction

Free Will as Indeterminacy
Reasons and Causal Explanations
The Agent-Self
Objections

Free Will as Self-Determination
Second-Order Desires and Valuing
Reactive Attitudes and Responsibility

Choice and Determinism

Conclusion

Bibliography

Introduction

“Two roads diverged in a yellow wood,/ And sorry I could not travel both/ And be one traveler …”^[1]

Free will is one of the most fundamental presuppositions that we make in our everyday lives. At the same time, free will belongs to the concepts that are most difficult to integrate into a scientific idea of how the world works. This contrast has haunted philosophers for centuries, and although it seems that almost everything has already been said about this topic, there has been an animated debate in recent decades.^[2] In this paper, I would like to consider two of the positions adopted in this debate.

At first, I would like to explore John Searle's interpretation of free will as a sequence of several moments of indeterminacy ("gaps") between the reasons for our actions and the actions themselves. I think the best way to understand his conception is to see it as an attempt to unite two different ideas about the relation between reasons and actions. On the one hand, the realist conception of Searle's philosophy presupposes that the reasons for our actions must have real causal power and are not only post facto justifications. On the other hand, Searle's understanding of rationality implies that reasons alone cannot be sufficient causes. In "Rationality in Action" Searle tries to bring both ideas together through the notion of an agent-self. I will argue that this attempt is problematic, not only because it leads our conception of free will towards a very unsatisfying choice between epiphenomenalism and natural indeterminism, but also because Searle cannot explain convincingly how the existence of indeterminacy in our brains could explain free will.

The second part of the paper is an attempt to develop an alternative notion of free will, relying on two influential articles by Harry Frankfurt and Peter Strawson. The basic idea is that when we are interested in free will, we are not primarily interested in causal origins, but in self-determination, which is also the main presupposition in the practice of holding people responsible. With this perspective in mind, I will finally return to the questions of the natural foundations of free will and its compatibility with a deterministic conception of nature.

In the paper, I will sometimes use the term "determinism", by which I mean, in Van Ingwaagen's words, "the thesis that there is at any instant exactly one possible future."^[3] It is important to see that determinism in this sense does not entail that we also know what the future will be.

Furthermore, in this paper I will assume that reasons and intentional states are valid explanations of actions. There are indeed radical reductionist positions which deny that intentional states have any real power in the explanation of actions; but it is questionable if these positions are meaningful.^[4] It seems to me that a concept of action purely in the form of neurophysiological states cannot explain what an action is, when it starts, when it ends, and how complex actions are possible.

Free Will as Indeterminacy

Reasons and Causal Explanations

One of the fundamental assumptions of Searle's theory of rationality is that the reasons, decisions and actions of rational agents are not simply parts of a causal chain, but are separated by different "gaps". This assumption is based on a psychological experience: if we look at our actions in retrospect, we do not think that our decisions are sufficient to cause actions. In our experience, having a reason for an action is not enough to actually make us act. If we look into the future, we experience these gaps as our ability to choose between several decisions and actions in the future.^[5]

Searle states that rational agents must have three of these gaps: the first gap must be between the reasons for an action and the making of the decision, the second between the decision and the onset of the action, and the third between the onset of the action and its continuation (in case of complex actions).^[6] All these gaps are psychologically expressed in the feeling that we as agents can intervene. For example, we feel that reasons do not compel us to make a certain decision, but that we still have to make the decision ourselves. Searle's thesis is that these gaps are what we mean when we talk about free will.

He emphasizes that this conception of ourselves as free agents is unavoidable, because we act, in Kant's words "under the idea of freedom". Thus, even a convinced determinist who sits in a restaurant cannot just wait for his desires to cause the decision for a certain meal, but he actually has to make a choice himself. Even his refusal to choose would be an exercise of freedom.^[7] For Searle, rational actions and free will are connected insofar as we must understand rational actions as the result of our choices, which excludes the idea of sufficient causation . Only irrational actions are sufficiently caused by our beliefs and desires. Actions which are caused by obsessions or addictions are examples of this kind of irrationality.^[8]

However, the gaps only exist as a psychological phenomenon, and it is not clear if they have a neurobiological foundation, that is, if the indeterminacy of the gaps corresponds to an indeterminacy in the structure of the brain. Thus, to ask if free will really exists is to ask what the ontology of the gap is. Searle differentiates between two possibilities^[9]: if our brain states really determine each other, as the standard account of nature suggests, our self-conception as agents with free will does not have any foundation in the nature of our brains and is therefore illusory. If there are, however, brain states which are not causally linked to previous brain states, free will would be a real natural phenomenon. However, this would also mean that we would have to assume the existence of indeterminacy in nature, i.e. there would have to be states of the world in general which are not sufficiently caused by previous states of the world; an assumption that is in conflict with the scientific conception of nature in classical physics.^[10]

Thus, the assumption that free will is a gap in our psychological experience which has to have a foundation in the structure of the brain leads to a problem. By asking if the psychological experience of indeterminism is based on any kind of indeterminism in nature, it seems that Searle has transformed the free will problem from a philosophical problem into an empirical question for natural scientists.

However, there are still two conceptual issues that must be explained if the notion of free will as the absence of sufficient causation should make sense: firstly, there must be a way to account for the difference between a decision by a rational agent on the one hand, and mere luck or chance on the other. If there are really no causally sufficient conditions to explain decisions and actions, it seems that the connection between reasons and actions can only be arbitrary.^[11] In fact, this argument is one of the main problems for every incompatibilist account of free will.

Second, if reasons are not sufficient causes for actions, what else are they? And how is it that we usually accept reason explanations as sufficient? One of the first possibilities that one could have in mind is to argue that reasons are merely rationalizations of actions and do not have causal power. But then the situation gets confusing, because it would seem as if actions would have reasons and causes. The big problem would then be to find a good answer to the question how reason and causal explanations are related to each other, for example if they can coexist or if they can be reduced to each other.

However, to differentiate between reasons and causes in this way is not acceptable for Searle, because his causal realism requires that our intentional states can really make things happen in the world. The deeper reason for that is that action as well as perception are both characterized by causal self-referentiality: to claim that an action is not really caused by our intentional states is in some way similar to claiming that a hallucination is a real experience of seeing something.^[12]

This leads Searle to the following assumption: "Reasons and rationality, in order to explain, must function causally (modulo the gap, of course)."^[13] If this were true, it would both explain what makes rational action different form pure luck, and how reasons can be sufficient explanations of actions. However, the obvious question is how this idea relates to the argument made above that reasons cannot be causes. Does "function causally" mean something different than "cause"? And what does "modulo the gap" mean?

The Agent-Self

Searle's central claim is that in order to explain how reasons can be sufficient explanations without being causes, we need to assume the existence of an agent-self. However, Searle explicitly rejects any "metaphysical" notion of the self in which the self appears as the bearer of some sort of personal identity through time.^[14] In Searle's account, the self must be an agent who is able to act on reasons; but agency alone is not enough: the self must also have certain cognitive abilities, such as a unified conscious field and the ability to have memory.^[15]

For Searle, this notion of the self makes sense of why reason explanations are usually accepted as sufficient explanations, although they do not cite causally sufficient factors: "The causal gap does not imply an explanatory gap."^[16] The question "why did you do that?" asks simply for the reason on which we acted, and not for the conditions that sufficiently caused us to act in a certain way.^[17]

It seems that Searle’s argument relies mainly on psychological evidence, for example in the intuition that actions and events are principally different or in the contrast between the passive experience of perception and action.^[18] In "Freedom and Neurobiology" he calls his own argument "transcendental" in Kant's sense, because it asks what the conditions of possibility of certain facts are: "I am claiming that the conditions of possibility of the adequacy of rational explanations is the existence of an irreducible self, a rational agent, capable of acting on reasons."^[19] In "Rationality in Action" Searle also talks about the "Background presupposition of the existence of rational selves"^[20].

[...]

^[1] Robert Frost, The Road not Taken

^[2] I found several collections with reprints of some classical essays very useful: Gary Watson (ed.), Free Will, Oxford 1982. Alfred Mele (ed.), The Philosophy of Action, Oxford 1997. Kane, Robert (ed.): The Oxford Handbook of Free Will, Oxford 2002. A good (though not wholly unbiased) source of articles is also Ted Honderich's "Determinism and Freedom Philosophy" website: http://www.ucl.ac.uk/%7Euctytho/dfwIntroIndex.htm.

^[3] Van Ingwagen, Peter: An Essay on Free Will, Oxford 1983, pp. 3.

^[4] Compare the discussion between: Malcolm, Norman: The Conveivability of Mechanism, in: Gary Watson (ed.), Free Will, Oxford 1982, pp. 127-149 and Kim, Jaegwon: Mechanism, Purpose, and Explanatory Exclusion, in: Alfred Mele (ed.), The Philosophy of Action, Oxford 1997, pp. 256-283.

^[5] Searle, John: Rationality in Action, Cambridge 2001, pp. 62.

^[6] Ibid., pp. 42.

^[7] Ibid., pp. 14.

^[8] Ibid., pp. 12.

^[9] Searle, John: Free Will as a Problem in Neurobiology, New York 2007, pp. 61 pass.

^[10] It seems, however, that even classical mechanics are not necessarily deterministic. For a good insight into the role of determinism in physical theories compare Hoefer, Carl: Art. "Causal Determinism", in: Stanford Encyclopedia of Philosophy, http://plato.stanford.edu/entries/determinism-causal/

^[11] Searle, Rationality in Action, pp. 80.

^[12] Searle, John: Intentionality. An Essay in the Philosophy of Mind, Cambridge 1983, pp 130.

^[13] Searle, Rationality in Action, pp. 113.

^[14] Ibid., pp. 76.

^[15] Ibid., pp. 84. For all the properties the self must have comp. pp. 95.

^[16] Ibid, pp. 85.

^[17] Ibid, pp. 86.

^[18] Ibid., pp. 74.

^[19] Searle, Freedom and Neurobiology, pp. 57.

^[20] Searle, Rationality in Action, pp. 84.