The Cogito Model
Cogito Model is the name that Bob Doyle, the Information Philosopher, gave to his two-stage model of free will in the 1970's, at the same time as he chose Ergo as a positive name for "negative" entropy, and Sum as the sum of all human knowledge. It was a gesture of appreciation to the father of modern philosophy, René Descartes, who first asserted that reason is the ultimate justification for knowledge. In his fifty years of philosophical research since then, Doyle has found another two dozen thinkers, some earlier - some later, who more or less independently argued for a two-stage model of free will. The Cogito Model has been refined to incorporate the best thinking of all those models and the best understanding of quantum physics and neuroscience. But this page preserves some of Doyle's earliest thinking - the Micro Mind and Macro Mind as id and superego, for example The Cogito Model of human freedom locates randomness (either ancient chance or modern quantum indeterminacy) in the mind, in a way that breaks the causal chain of strict physical determinism, while doing no harm to responsibility. The Cogito Model combines indeterminacy - first microscopic quantum randomness
and unpredictability, then "adequate" or statistical determinism and macroscopic predictability,
in a temporal sequence that creates new information. Important elements of the model have been proposed by many philosophers since Aristotle, the first indeterminist. A number of modern philosophers and scientists, starting with William James, have proposed similar two-stage models of free will. But none of them has been able to locate the randomness so as to make free will "intelligible," as the libertarian "free-willist" philosopher Robert Kane puts it. The insoluble problem for previous two-stage models has been to explain how a random event in the brain can be timed and located - perfectly synchronized! - so as to be relevant to a specific decision. The answer is it cannot be, for the simple reason that quantum events are totally unpredictable. The Cogito solution is not single random events, one per decision, but many random events in the brain as a result of ever-present noise, both quantum and thermal noise, that is inherent in any information storage and communication system. The mind, like all biological systems, has evolved in the presence of constant noise and is able to ignore that noise, unless the noise provides a significant competitive advantage, which it clearly does as the basis for freedom and creativity. Let's see how randomness in the Cogito Model is never the direct cause of our decisions. Decisions are always adequately determined. We assume that there are always many contributing causes for any event, and in particular for a mental decision. In the Newell-Simon "Blackboard" model and Bernard Baars' "Theater of Consciousness" and "Global Workspace" models, there are many competing possibilities for our next thought or action. Each of these possibilities is the result of a sequence of events that goes back in an assumed causal chain until its beginning in an uncaused event. Aristotle called this original event an arche (ἀρχῆ), one whose major contributing cause (or causes) was itself uncaused, e.g., it involved quantum indeterminacy. What this means is that tracing any particular sequence of events back in time will come to one event - a "starting point" or "fresh start" (Aristotle calls it an origin or arche (ἀρκῆ)) - whose major contributing cause (or causes) was itself uncaused, in that it involved quantum indeterminacy.
Whether a particular thing happens, says Aristotle, may depend on a series of causes that
"goes back to some starting-point, which does not go back to something else. This, therefore, will be the starting-point of the fortuitous, and nothing else is the cause of its generation."
We can thus in principle assign times, or ages, to the starting points of the contributing causes of a decision. Some of these may in fact go back before the birth of an agent, hereditary causes for example. To the extent that such causes adequately determine an action, we can understand why hard determinists think that the agent has no control over such actions. (Of course if we can opt out of the action at the last moment, we retain a kind of control.)
Other contributing causes may be traceable back to environmental and developmental events, perhaps education, perhaps simply life experiences, that were "character-forming" events. These and hereditary causes would be present in the mind of the agent as fixed habits, with a very high probability of "adequately determining" the agent's actions in many situations.
But other contributing causes of a specific action may have been undetermined up to the very near past, even fractions of a second before an important decision. The causal chains for these contributing causes originate in the noisy brain. They include the free generation of new alternative possibilities for thought or action during the agent's deliberations. They fit Aristotle's criteria for causes that "depend on us" (ἐφ' ἡμῖν) and originate "within us" (ἐv ἡμῖν).
These alternatives are likely generated from our internal knowledge of practical possibilities based on our past experience. Those that are handed up for consideration to Baars' "executive function" may be filtered to some extent by unconscious processes to be "within reason." They likely consist of random variations of past actions we have willed many times in the past.
Note that the evaluation and selection of one of these possibilities by the will is as deterministic and causal a process as anything that a determinist or compatibilist could ask for, consistent with our current knowledge of the physical world.
Why have philosophers been unable for millennia to accept the common sense view that humans are free? Partly because their logic and language preoccupation makes them say that either determinism or indeterminism is true, and the other must be false. This is the standard (but flawed) argument against free will. Our physical world includes both, although the determinism we have is only an adequate description for large objects. So any intelligible explanation for free will must include both indeterminism and adequate determinism.
Micro Mind and Macro Mind
Imagine a Micro Mind with a randomly assembled "agenda" of possible things to say or to do. These are drawn from our memory of past thoughts and actions, but randomly varied by unpredictable negations, associations of a part of one idea with a part or all of another, and by substitutions of words, images, feelings, and actions drawn from our experience. In information communication terms, there is cross-talk and noise in our neural circuitry.
In a "content-addressable" information model, memories are stored based on their content - typically bundles of simultaneous images, sounds, smells, feelings, etc. So a new experience is likely to be stored in neural pathways alongside closely related past experiences. And a fresh experience, or active thinking about an experience that presents a decision problem, is likely to activate nearby brain circuits, ones that have strong associations with our current circumstances. These are likely to begin firing randomly, to provide unpredictable raw material for actionable possibilities.
The strong feeling that sometimes "we don't know what we think until we hear what we say" reflects our capability for original and creative thoughts, different from anything we have consciously learned. Something as simple as substituting a synonymous word, or more complex replacements with associated words (metonyms) or wild leaps of fancy (metaphor) are examples of building unpredictable thoughts. Picturing ourselves doing something we have seen others do, from "monkey see, monkey do" childhood mimicry to adult imitations, is a source for action items on the agenda, with the random element as simple as if and when we choose to do them.
The etymology of cogito is Latin co-agitare, to shake together. Why do we need quantum uncertainty involved in the shaking together of our agenda items? Will neuroscientists ever find information structures in the brain to generate our random agenda, structures small enough to be susceptible to microscopic quantum phenomena? Some fanciful mechanisms include a random number generator like those in computer programs. Others imagine a microscopic event like the nuclear decay in Schrödinger's diabolical thought experiment with a cat, plus the amplifier circuitry needed to magnify the event to the macroscopic level where it is naively thought to help make the decision. But chance cannot be the direct cause of our actions.
Nothing physically localized is likely to be found. The randomness of the Micro Mind is simply the result of ever-present noise, both thermal and quantum noise, that is inherent in any information storage and communication system.
Constant, ever-present noise removes an important technical objection. Critics of the Epicurean swerve of the atoms asked when and where and how would the random event occur?. The Cogito model randomly generates contextually appropriate alternative possibilities at all times.
The Cogito model is not a mechanism. It is a process, and information philosophy is a process philosophy.
Quantum uncertainty adds a "causa sui," an uncaused or self-caused cause, in the causal chain. But it need not directly determine the decision of the macroscopic will or the fully determined resulting action which is consistent with character and values.
Some argue that brain structures are too large to be affected at all by quantum events. But there is little doubt that the brain has evolved to the point where it can access quantum phenomena. The evolutionary advantage for the mind is freedom and creativity. Biophysics tells us the eye can detect a single quantum of light (photon), and the nose can smell a single molecule.
If the Micro Mind is a random generator of frequently outlandish and absurd possibilities, the complementary Macro Mind is a macroscopic structure so large that quantum effects are negligible. It is the critical apparatus that makes decisions based on our character and values.
Information about our character and values is probably stored in the same noise-susceptible neural circuits of our brain, in our memory, so Macro Mind and Micro Mind are not necessarily in different locations in the brain. Instead, they are probably the consequence of different information processing methods. The Macro Mind must suppress the noise when it makes an adequately determined decision.
The Macro Mind has very likely evolved to add enough redundancy, perhaps even error detection and correction, to reduce the noise to levels required for an adequate determinism. Our decisions are then in principle predictable, given knowledge of all our past actions and given the randomly generated possibilities in the instant before decision. However, only we know the contents of our minds, and they exist only within our minds. Thus we can feel fully responsible for our choices, morally and legally.
The model accounts not just for freedom but for creativity, original thoughts and ideas never before expressed. Unique and new information comes into the world with each new thought and action.
Biologists will note that the Micro Mind corresponds to random variation in the gene pool (often the direct result of quantum accidents). The Macro Mind corresponds to natural selection by highly determined organisms. See the biology chapter for other examples of random generation followed by adequately determined selection, like the immune system and protein/enzyme factories. Karl Popper may have been the first to point this out.
Psychologists will see the resemblance of Micro Mind and Macro Mind to the Freudian id and super-ego (das Ess und das Über-ich).
The model accounts quantitatively for the concept of wisdom. The greater the amount of knowledge and experience, the more likely that the random agenda will contain more useful and "intelligent" thoughts and actions as alternative possibilities. It also implies that an educated mind is "more free" because it can generate a wider agenda and options for action. It suggests that "narrow" and "closed" minds may simply be lacking the capabilities of the Micro Mind. And if the Macro Mind were weak, it might point to the high correlation between creativity and madness suggested by a Micro Mind out of control.
Philosophers of Mind, whether hard determinist or compatibilist, should recognize this Macro Mind as everything they say is needed to make a carefully reasoned free choice. But now choices include self-generated random possibilities for thought and action that no external agent can predict. Thus the choice of the will and the resulting willed action are unpredictable. The origin of the chosen causal chain is entirely within the agent, a condition noted first by Aristotle for voluntary action, his ἐν ἡμῖν ("in us").
The combination of microscopic randomness and macroscopic determinism in our Cogito model for human freedom means it is both unpredictable and yet fully responsible for its willed actions. Chance never leads directly to - never directly "causes" - an action.
Chance only provides the variety of alternative possibilities, each the possible start of a new causal chain, from which the deterministic judgment can choose an alternative that is consistent with its character and values. Our will is adequately determined and in control of our actions.
In summary, we distinguish six increasingly sophisticated ideas about the role of chance and indeterminism in the question of free will. Many libertarians have accepted the first two. Determinist and compatibilist critics of free will make the third their central attack on chance. But very few thinkers appear to have considered all six essential requirements for chance to contribute to our Cogito model of libertarian free will, specifically the last three requirements - that chance must be present but suppressible at will.
In our Cogito model, "Free Will" combines two distinct concepts. Free is the chance and randomness of the Micro Mind. Will is the adequately determined choice of the Macro Mind. And these occur in a temporal sequence. Since the chance suggestions for alternative possibilities appear first in the theater of consciousness (though they are largely unconscious and competing for attention), the delay before a conscious choice by the frontal lobe could easily account for the results of Benjamin Libet's experiments.
Compatibilists and Determinists were right about the Will,
The Temporal Sequence of Freedom and Determination
Free Will is best understood as a complex idea combining two antagonistic concepts - freedom and determination.
Many philosophers have called free will "unintelligible" because of this internal contradiction and the presumed simultaneity and identity of free and will. Specifically, they mistakenly have assumed that "free" is a time-independent adjective modifying "will." And they have often taken "free" pejoratively to mean "random."
A careful examination of ordinary language usage shows that free will is actually a temporal sequence of two opposing concepts - first "free" and then "will."
First comes the consideration of alternative possibilities, which are generated unpredictably by acausal events (simply noise in neural network communications). This free creation of possible thoughts and actions allows one to feel "I can do otherwise."
Next comes de-liberation and determination by the will, the unfreeing of possibilities into actuality, the decision that directs the tongue or body to speak or act.
After the deliberation of the will, the true sentence "I can do otherwise" can be changed to the past tense and remain true as a "hard fact" in the "fixed past," and written "I could have done otherwise."
Thus we have the temporal sequence which William James saw so clearly a century ago, with chance in a present time of random alternatives, leading to a choice which grants consent to one possibility and transforms an equivocal future into an unalterable and simple past. Free undetermined alternatives are followed by willed, determined choices. As John Locke knew more than three hundred years ago, "free" is an adjective that describes not the will, but the human mind. Just as "free" needs to be separated from "will," we think "moral" should be separated from "responsibility." Furthermore "free will" should be separated from "moral responsibility" and "moral responsibility" should be separated from "retributive punishment" and vengeance. See our argument for this critical conceptual analysis of concepts in the free will debates.
A More Detailed LookGiven the "laws of nature" and the "fixed past" just before a decision, philosophers wonder how a free agent can have any possible alternatives. This is partly because they imagine a timeline for the decision that shrinks the decision process to a single moment.
Thoughts are FreeOur thoughts are free and often appear to come to us. Our actions are adequately determined for moral responsibility and appear to come from us. They are up to us (Aristotle's ἐφ' ἡμῖν). What then are the sources of alternative possibilities? To what extent are they our creations? We can distinguish three important sources, all of them capable of producing indeterministic options for thoughts and actions. The first source is the external world that arrives through our perceptions. It is perhaps the major driving force in our lives, constantly requiring our conscious attention. Indeed, consciousness can be understood in large part as the exchange of actionable information between organism and environment. Although the indeterminisitic origin of such ideas is outside us, we can take full responsibility for them if they influence our adequately determined willed actions. The second source of options is other persons. The unique human ability to communicate information means that alternative possibilities for our actions are being generated by our reactions to other minds. Peter Strawson's reactive attitudes come to mind. Finally, and most importantly, our Micro Mind generates possibilities internally. These are the possibilities that truly originate within us (Aristotle's ἐν ἡμῖν). Note that the sources of random options not only need not be internal, they need not be contemporaneous with the current decision, as long as they come to mind as alternatives. They may have been generated at much earlier times in the agent's life, and only now get reconsidered and perhaps get acted upon.
How does the Cogito Model Compare to Other Mind Models?
DeterminismDeterminists think everything that happens is strictly caused by the prior states of the world, so there is no free will or moral responsibility. Free will is an illusion. Free will is impossible.
CompatibilismCompatibilists say it is freedom enough if we are not externally coerced. Since our will is one of the causes in the chain of causes, that is enough to make us morally responsible for our actions.
LibertarianismLibertarians demand that there are breaks in the causal chain of determinism. These breaks can be random events in the past, random events in our deliberations immediately before decisions, or randomness in our decisions themselves. These are called "event-causal" views. The breaks could also be caused by metaphysical powers in the agent that can start new causal chains. These are called "agent-causal or "non-causal" views. An extreme form of "substance dualism" makes the mind a different substance from the body.
The Cogito Model in the Spectrum of Free Will PositionsThe Cogito Model lies between and combines elements of Determinism and Agent-Causal Libertarianism The Cogito Model is less libertarian than typical event-causal Libertarians and more libertarian than some recent event-causal models proposed by Compatibilists that attempt to "give Libertarians what they want." The Cogito Model is in part an "event-causal" view that locates the breaks in the causal chain in our deliberations, These include the internal uncaused generation of new possibilities as well as external random sensory inputs. It is less radically libertarian and more determinist than the typical models of Libertarians Robert Kane, Laura Waddell Ekstrom, and Mark Balaguer which locate randomness in the decisions themselves. The Cogito Model is compatibilist in the sense that it is compatible with "adequate" determinism. It is very similar to the two-stage models of Daniel Dennett and Alfred Mele. But unlike Dennett, the model needs quantum randomness and not simply computational "pseudo-randomness" to generate alternative possibilities. And unlike Mele, we believe that science has shown indeterminism to be true and determinism to be false. Mele remains an agnostic on these important questions. Again, beyond the Dennett and Mele models, the Cogito Model proposes a specific process that avoids the single "quantum event in the brain" that gets amplified perfectly in time with our thought processes to help with free will. There are billions of quantum events in the brain every second. The miracle of the mind is that it can manage the resulting noise, averaging over these events when it needs to, utilizing them when it wants to. Because the agent is actively controlling the process of deliberation up to the instant of the determining decision at the 'moment of choice,' the Cogito Model shares much with agent-causal views, without being metaphysical. The "free" stage of the Cogito Model depends on thermal and quantal noise in the neural circuitry of the brain. This noise introduces errors in the storage and retrieval of information, noise that may be helpful in generating alternative possibilities for action. The "will" stage of the Cogito Model suppresses this noise for the adequately determined process of evaluation and decision, unless the will is satisfied with a random choice in special cases of the liberty of indifference. The Cogito Model is compatible with indeterminism suitably located and determinism appropriately limited. It is thus "doubly compatible" with indeterminism and "adequate" determinism. This suggests what we call a Comprehensive Compatibilism. The Cogito Model is arguably closer to the common sense or "folk" view of free will than any other free will model.
Objections to the Cogito ModelThe earliest objections were the concerns of some of the inventors of two-stage models themselves. Mostly they could not see how to reconcile the randomness of indeterminism with the determinism required for responsibility. They also tended to be metaphysical dualists, so they did not have a purely physical model for free will. Arthur Holly Compton adhered to a view that human freedom might only be visible from the inside (subjectively), that from the outside a person would be seen (objectively) as deterministic. This was a variation on Neils Bohr's dualist complementarity principle, which was popular among physicists at the time. Karl Popper, in his collaborations with the neurobiologist John Eccles, wanted the will to involve a metaphysical interaction between the mind (or soul) and the body. This was another form of dualism. Later (1977), Popper endorsed the idea of a two-stage model with quantum indeterminacy in the first stage, and a lawful determined selection process similar to Darwinian evoltuion. Henry Margenau wrestled with his mentor Ernst Cassirer's views on determinism and indeterminism in physics. Cassirer also had strong Kantian dualism tendencies, but in the end he insisted that only determinism could provide the causality needed as a basis for science. Margenau reluctantly accepted indeterminism as the "first step" in an explanation of human freedom and possibly providing insight into ethical problems.
Daniel Dennett's Objections to his own Valerian modelIn 1978, Daniel Dennett proposed a two-stage model that would "give the libertarians what they want." But he had serious reservations about his "Valerian" model, most important that he could find no place in it for quantum indeterminism. Dennett's model for decision making started with elements from Henri Poincaré random combinations model (via Jacques Hadamard and the poet Paul Valery, at the 1936 Synthese conference in Paris exploring creativity). Dennett mentioned the amplification of a quantum event in the brain, which was first suggested by Arthur Holly Compton in 1931. Dennett had also read Karl Popper, who had criticized Compton's "massive switch amplifier." He knew Popper's analogy of free will with natural selection as a two-stage process. Dennett's decision-making model was a variation of computer scientist Herbert Simon's "generate and test" two-stage model for computer problem solving. Dennett made an excellent case for his model as something that libertarians should want. Sadly, no libertarian saw the power in Dennett's two-stage model. Because Dennett saw very clearly what was good about the model for Libertarians, he also could see what they might not accept. Dennett knew that some libertarians insisted on indeterministic quantum events in the brain, but he could not understand the place for a quantum event, how exactly and when and where a quantum event in the brain could be amplified to help with decision making and not harm our control and responsibility for our actions. As a determinist, Dennett said that a model with pseudo-random number generation in the first stage would be all that is needed. He found no value in adding true quantum randomness.
Robert Kane's Objections to Dennett's two-stage modelBob Kane had independently developed the two-stage model before Dennett published Brainstorms. He had read the same sources (Compton and Popper), but he thought that something more was needed. Basically, Kane felt that at the completion of the first stage in the model, when all the random considerations have been generated, there is a finite time, however small, during which the model assumes that the willed decision, the choice between alternative possibilities, is determined. This is unacceptable for an incompatibilist. Libertarian free will requires that the decision not be determined, even by the agents desires and beliefs. It must remain undetermined up to the moment of choice. It is determined by the choice, says Kane. The agent does not have complete control over the random considerations that get generated. The agent can decide to stop generating new possibilities. And if evaluation finds none satisfactory, can go back and generate more. But after the last new random option is generated, and during that time, however small, before the decision is made, Kane is concerned that the choice is already determined by the agent's character, reasons, motives, etc. Kane is correct. In our Cogito model, the decision could be reliably (though not perfectly) predicted by a super-psychiatrist who knew everything about the agent and was aware of all the alternative possibilities. This is because the second ("will") stage evaluation and decision process is indeed adequately determined. We agree with Kane that the second stage is determined, in the sense of adequately determined, but note that it is in no way pre-determined. And Kane agrees that, before the first stage of the two stage model, the decision is not determined. It is at that time undetermined. Now Kane has found a way to maintain that in some cases, the agent's decision may not be determined by anything other than the agent's choice, which can be rational (made for properly evaluated reasons), but nevertheless might have been otherwise and yet be equally rational. He calls this "dual (or plural) rational control. We shall see that this is an acceptable extension of the Cogito model. Let's see how it works. To find a way around the determined second stage, without invoking metaphysical agent-causality, Kane adds event-causal randomness in the decision itself. Randolph Clarke calls such randomness "centered" in the decision, as opposed to chance located earlier in the "deliberative" stage (our first "free" stage). Kane calls this The Indeterminist Condition:
"the agent should be able to act and act otherwise (choose different possible futures), given the same past circumstances and laws of nature." ( A Contemporary Introduction to Free Will, 2005, p.38)In the Cogito model, the agent has the capacity to "act otherwise" at the start of the first, "free" stage of the two-stage model. At the beginning of the second ("will") stage, the choice and the act has been adequately determined. But now Kane adds "something more" to the two-stage model. There are times when the deliberation and evaluation process is not at all simple and straightforward. The agent may be seriously conflicted, especially in difficult moral decisions, which Kane says contribute to what he calls Self-forming Actions (SFAs). In such cases, the agent has developed excellent reasons for more than one option. This conflict requires extra effort on the part of the agent to make the decision, which Kane says may generate noise in the brain's neural circuitry . This noise may make the decision random and indeterminate, although it selects from among options that are all defended by reasons. Kane deftly sidesteps the charge of critics who the agent cannot be responsible for any decision involving randomness. In Kane's model, the agent can claim responsibility however the "torn" decision is made. Kane's followers today, Laura Waddell Ekstrom and Mark Balaguer, for example, continue to argue for randomness directly in the decision as the only way to resolve "torn decisions." Here is Ekstrom's position, for example:
Consider an agent whose act is, in such a sense, "libertarian free." Now a duplicate agent in exactly similar circumstances governed by the same natural laws and subject to the same occurrence of considerations at the same points in the deliberative process will form exactly the same judgment concerning the best thing to do and will act accordingly. But then, given the consideration pattern that occurs (but might not have), there is no "wiggle room" for the agent in forming an evaluative judgment — it simply falls out, of necessity, from the consideration pattern. Hence such an account does not leave sufficient room for free agency.
Richard Double's Objections to Kane's "dual rational control."Kane's position is not without its critics. Double's objections are mostly directed at Kane's efforts to establish moral responsibility. Double develops challenges to three of Kane's requirements: the ability to have chosen otherwise, control, and rationality. Double noted that Dennett's Valerian models introduce indeterminism in the early stages of deliberation, before the decision itself.
He therefore calls Kane's views "Non-Valerian." These allow indeterminism in the decision process itself, which means that chance is sometimes the direct cause of actions, which Double feels jeopardizes agent control.
The last is his own theory, which he calls "Delay Libertarianism." The main idea is to recognize that free will is a process that takes place over a period of time. This gives Double the opportunity to locate the indeterminism in a delay between deliberations and resultant decisions.
Double notes that the deliberations "set the stage" for whatever decision will be made - if any decision is made. He does not show how delayed indeterminism can resolve the randomness objection.
Double recognizes that the act of the will might be simply to avoid a decision, and send the problem back for more deliberations, which could involve generating more alternative possibilities, as in our Cogito Model.
But in the end, says Double, delay libertarianism fails, for the same reason as the others, the dual rational control condition.Dual rational control is Kane's claim that the agent can do otherwise (randomly) and have the alternative (dual) action be just as rational and demonstrate just as much control as the original action. Double is right to reject this view, but unfortunately he rejects all libertarianism, titling his book The Non-Reality of Free Will.
Alfred Mele's Reservations about his "Modest Libertarianism."Mele's "Modest Libertianism" is essentially the same as Dennett's "Valerian" model:
The modest indeterminism at issue allows agents ample control over their deliberation. Suppose a belief, hypothesis, or desire that is indirectly relevant to a deliberator's present practical question comes to mind during deliberation but was not deterministically caused to do so. Presumably, a normal agent would be able to assess this consideration. And upon reflection might rationally reject the belief as unwarranted, rationally judge that the hypothesis does not merit investigation, or rationally decide that the desire should be given little or no weight in his deliberation. Alternatively reflection might rationally lead him to retain the belief, to pursue the hypothesis to give the desire significant weight. That a consideration is indeterministically caused to come to mind does not entail that the agent has no control over how he responds to it. Considerations that are indeterministically caused to come to mind (like considerations that are deterministically caused to come to mind) are nothing more than input to deliberation. Their coming to mind has at most an indirect effect on what the agent decides, an effect that is mediated by the agent's assessment of them. They do not settle matters. Moreover, not only do agents have the opportunity to assess these considerations, but they also have the opportunity to search for additional relevant considerations before they decide, thereby increasing the probability that other relevant considerations will be indeterministically caused to come to mind. They have, then, at least sometimes, the opportunity to counteract instances of bad luck — for example, an indeterministically caused coming to mind of a misleading consideration or, a chance failure to notice a relevant consideration. And given a suitable indeterminism regarding what comes to mind in an assessment process, there are causally open alternative possibilities for the conclusion or outcome of that process. Compatibilists who hold that we act freely even when we are not in control of what happens at certain specific junctures in the process leading to action are in no position to hold that an indeterministic agent's lacking control at the same junctures precludes free action. And, again, real human beings are not in control of the coming to mind of everything that comes to mind during typical processes of deliberation. If this lack of perfect proximal control does not preclude its being the case that free actions sometimes issue from typical deliberation on the assumption that we are deterministic agents, it also does not preclude this on the assumption that we are indeterministic agents. Now, even if garden-variety compatibilists can be led to see that the problem of luck is surmountable by a libertarian, how are theorists of other kinds likely to respond to the libertarian position that I have been sketching? There are, of course, philosophers who contend that moral responsibility and freedom are illusions and that we lack these properties whether our universe is deterministic or indeterministic — for example, Richard Double and Galen Strawson. Modest libertarians can also anticipate trouble from traditional libertarians, who want more than the modest indeterminism that I have described can offer. Clarke, who has done as much as anyone to develop an agent-causal libertarian view, criticizes event-causal libertarianism on the grounds that it adds no "positive" power of control to compatibilist control but simply places compatibilist control in an indeterministic setting. Of course, given that combining compatibilist control with indeterminism in a certain psychological sphere was my explicit strategy in constructing a modest libertarian position, I do not see this as an objection. In any case, traditional libertarians need to show that what they want is coherent.Mele is correct that his model will not satisfy Libertarians wanting more, whether "agent-causal" libertarians like Timothy O'Connor or "event-causal" libertarians Robert Kane wanting randomness in their decisions.
Randolph Clarke's Objections to Dennett, Mele, Ekstrom, and Kane.Clarke defines new terms in his Libertarian Accounts of Free Will for Double's "Valerian" and "Non-Valerian." He calls Dennett's model "deliberative," since randomness internal to the mind is limited to the deliberations. And he calls Kane's model "centered," by which he means that Kane's (quantum) randomness is in the center of the decision itself. Clarke accepts the Kane and Ekstrom views that if the agent's decision simply results from indeterministic events in the deliberation phase that that could not be what he calls "directly free." Clarke calls this deliberative freedom "indirect." "Indirectly free" is a reasonable description for our Cogito Model, which limits indeterminism to the "free" deliberation stage and has a limited but "adequate" determinism in the "will" stage. Although Clarke says that a "centered event-causal libertarian view provides a conceptually adequate account of free will," he doubts that it can provide for moral responsibility. He says that
An event-causal libertarian view secures ultimate control, which no compatibilist account provides. But the secured ultimacy is wholly negative: it is just (on a centered view) a matter of the absence of any determining cause of a directly free action. The active control that is exercised on such a view is just the same as that exercised on an event-causal compatibilist account.It is a bit puzzling to see how the active control of a libertarian decision based on quantum randomness is "just the same as that exercised" on a compatibilist account, unless it means, as Double argued, no control at all. So it may be worth quoting Clarke at length.
Dennett requires only that the coming to mind of certain beliefs be undetermined; Mele maintains that (in combination with the satisfaction of compatibilist requirements) this would suffice, as would the undetermined coming to mind of certain desires. Likewise, on Ekstrom's view, we have undetermined actions — the formations of preferences — among the causes of free decisions. But she does not require that these preference-formations either be or result from free actions. Nor can she require this. Any free action, she holds, must be preceded by a preference-formation. An infinite regress would be generated if these preference-formations had to either be or result from free actions. And a similar regress would result if Dennett or Mele required that the undetermined comings-to-mind, attendings, or makings of judgments that figure in their accounts had to either be or result from free actions. Thus, given the basic features of these views, all three must allow that an action can be free even if it is causally determined and none of its causes, direct or indirect, is a free action by that agent. Setting aside the authors currently under discussion, it appears that all libertarians disallow such a thing. What might be the basis for this virtual unanimity? When an agent acts with direct freedom — freedom that is not derived from the freedom of any earlier action— she is able to do other than what she, in fact, does. Incompatibilists (libertarians included) maintain that, if events prior to one's birth (indirectly) causally determine all of one's actions, then one is never able to do other than perform the actions that one actually performs, for one is never able to prevent either those earlier events or the obtaining of the laws of nature.Clarke now claims that even prior events thought up freely by the agent during deliberations will "determine" the agent's decision. This is roughly what the Cogito Model claims. After indeterminism in the "free" deliberation stage, we need "adequate" determinism in the "will" stage to insure that our actions are consistent with our character and values (including Kane's SFAs), with our habits and (Ekstrom's) preferences, and with our current feelings and desires. Clarke oddly attempts to equate events prior to our births with events in our deliberations, claiming that they are equally determinist. He says,
If this is correct, then a time-indexed version of the same claim is correct, too. If events that have occurred by time t causally determine some subsequent action, then the agent is not able at t to do other than perform that action, for one is not able at t to prevent either events that have occurred by t or the obtaining of the laws of nature. An incompatibilist will judge, then, that, on Dennett's and Mele's views, it is allowed that once the agent has made an evaluative judgment, she is not able to do other than make the decision that she will, in fact, make, and that, on Ekstrom's view, it is allowed that once the preference is formed, again the agent is not able to avoid making the decision that she will, in fact, make. If direct freedom requires that, until an action is performed, the agent be able to do otherwise, then these views do not secure the direct freedom of such decisions. Mele confronts this line of thinking head-on. Some libertarians, he acknowledges, do hold that a decision is directly free only if, until it is made, the agent is able to do other than make that decision, where this is taken to require that, until the action occurs, there is a chance that it will not occur. But such a position, Mele charges, is "mere dogmatism" (1995a: 218). It generates the problem of control that he (along with Dennett and Ekstrom) seeks to evade, and hence libertarians would do well to reject this position. There is, however, a decisive reason for libertarians not to reject this position, a reason that stems from the common belief — one held by compatibilists and incompatibilists alike — that, in acting freely, agents make a difference, by exercises of active control, to how things go. The difference is made, on this common conception, in the performance of a directly free action itself, not in the occurrence of some event prior to the action, even if that prior event is an agent-involving occurrence causation of the action by which importantly connects the agent, as a person, to her action. On a libertarian understanding of this difference-making, some things that happen had a chance of not happening, and some things that do not happen had a chance of happening, and in performing directly free actions, agents make the difference. If an agent is, in the very performance of a free action, to make a difference in this libertarian way, then that action itself must not be causally determined by its immediate antecedents. In order to secure this libertarian variety of difference-making, an account must locate openness and freedom-level active control in the same event — the free action itself — rather separate these two as do deliberative libertarian views. On the views of Dennett, Ekstrom, and Mele, agents might be said to make a difference between what happens but might not have and what does not happen but might have, but such a difference is made in the occurrence of something nonactive or unfree prior to the action that is said to be free, not in the performance of the allegedly free action itself. Failure to secure for directly free actions this libertarian variety of difference-making constitutes a fundamental inadequacy of deliberative libertarian accounts of free action.To see that the Cogito Model allows the agent to make a real difference, we need only extend the process of decision to include everything from the start of free deliberations to the moment of willed choice The agent is justified saying "I could have done otherwise," "This action was up to me," and "I am the originator of my actions and the author of my life." Clarke goes on to consider his "centered" event-causal view, and initially claims that it provides an adequate account of free will, but his "adequate" is damning with faint praise.
If merely narrow incompatibilism is correct, then an unadorned, centered event-causal libertarian view provides a conceptually adequate account of free will. Such a view provides adequately for fully rational free action and for the rational explanation — simple, as well as contrastive — of free action. The indeterminism required by such a view does not diminish the active control that is exercised when one acts. Given incompatibilism of this variety, a libertarian account of this type secures both the openness of alternatives and the exercise of active control that are required for free will. It is thus unnecessary to restrict indeterminism, as deliberative accounts do, to locations earlier in the processes leading to free actions. Indeed, so restricting indeterminism undermines the adequacy of an event-causal view. Any adequate libertarian account must locate the openness of alternatives and freedom-level active control in the same event — in a directly free action itself. For this reason, an adequate event-causal view must require that a directly free action be nondeterministically caused by its immediate causal antecedents If, on the other hand, broad incompatibilism is correct, then no event-causal account is adequate. An event-causal libertarian view secures ultimate control, which no compatibilist account provides. But the secured ultimacy is wholly negative: it is just (on a centered view) a matter of the absence of any determining cause of a directly free action. The active control that is exercised on such a view is just the same as that exercised on an event-causal compatibilist account. This sort of libertarian view fails to secure the agent's exercise of any further positive powers to causally influence which of the alternative courses of events that are open will become actual. For this reason, if moral responsibility is precluded by determinism, the freedom required for responsibility is not secured by any event-causal libertarian account. (ibid., pp.219-20)So for Clarke all libertarian accounts fail if broad incompatiblism is true(viz. if determinism is incompatible with moral responsibility, if semicompatibilism is not true). moral responsibility. Some critics have mistakenly made it an objection to libertarian free will. Since the world contains irreducible chance, many unintended consequences of our actions are out of our control. Unfortunately, much of what happens in the real world contains a good deal of luck. Luck gives rise to many of the moral dilemmas that lead to moral skepticism. Whether determinist, compatibilist, semicompatibilist, or libertarian, it seems unreasonable to hold persons responsible for the unintended and unforeseeable consequences of their actions, good or bad. In many moral and legal systems, it the person's intentions that matter first and foremost. Nevertheless, we are often held responsible for actions that were intended as good, but that had bad consequences. Similarly, we occasionally are praised for actions that were either neutral or possibly blameworthy, but which had good consequences. Some thinkers are critical of any free will model that involves chance, because the apparent randomness of decisions would make such free will unintelligible. They say our actions would be a matter of luck. This is the Luck Objection to free will.
Moral judgment of a person is judgment not of what happens to him, but of him. It does not say merely that a certain event or state of affairs is fortunate or unfortunate or even terrible. It is not an evaluation of a state of the world, or of an individual as part of the world. We are not thinking just that it would be better if he were different, or did not exist, or had not done some of the things he has done. We are judging him, rather than his existence or characteristics. The effect of concentrating on the influence of what is not under his control is to make this responsible self seem to disappear, swallowed up by the order of mere events.
I entirely agree with [Nagel] that the involvement of morality with luck is not something that can simply be accepted without calling our moral conceptions into question. That was part of my original point; I have tried to state it more directly in the present version of this paper. A difference between Nagel and myself is that I am more sceptical about our moral conceptions than he is. Scepticism about the freedom of morality from luck cannot leave the concept of morality where it was, any more than it can remain undisturbed by scepticism about the very closely related image we have of there being a moral order, within which our actions have a significance which may not be accorded to them by mere social recognition. These forms of scepticism will leave us with a concept of morality, but one less important, certainly, than ours is usually taken to be; and that will not be ours, since one thing that is particularly important about ours is how important it is taken to be.
Agents' control is the yardstick by which the bearing of luck on their freedom and moral responsibility is measured. When luck (good or bad) is problematic, that is because it seems significantly to impede agents' control over themselves or to highlight important gaps or shortcomings in such control. It may seem that to the extent that it is causally open whether or not, for example, an agent intends in accordance with his considered judgment about what it is best to do, he lacks some control over what he intends, and it may be claimed that a positive deterministic connection between considered best judgment and intention would be more conducive to freedom and moral responsibility. This last claim will be regarded as a nonstarter by anyone who holds that freedom and moral responsibility require agential control and that determinism is incompatible with such control. Sometimes it is claimed that agents do not control anything at all if determinism is true. That claim is false. As soon as any agent...judges it best to A, objective probabilities for the various decisions open to the agent are set, and the probability of a decision to A is very high. Larger probabilities get a correspondingly larger segment of a tiny indeterministic neural roulette wheel in the agent's head than do smaller probabilities. A tiny neural ball bounces along the wheel; its landing in a particular segment is the agent's making the corresponding decision. When the ball lands in the segment for a decision to A, its doing so is not just a matter of luck. After all, the design is such that the probability of that happening is very high. But the ball's landing there is partly a matter of luck. All libertarians who hold that A's being a free action depends on its being the case that, at the time, the agent was able to do otherwise freely then should tell us what it could possibly be about an agent who freely A-ed at t in virtue of which it is true that, in another world with the same past and laws of nature, he freely does something else at t. Of course, they can say that the answer is "free will." But what they need to explain then is how free will, as they understand it, can be a feature of agents — or, more fully, how this can be so where free will, on their account of it, really does answer the question. To do this, of course, they must provide an account of free will — one that can be tested for adequacy in this connection.Free will is a prerequisite for responsibility. Whether a free action involves moral responsibility is a question for the ethicists. But in any case, to the extent that luck is involved in an agent's free actions, that is a problem for moral responsibility, which we can separate from the problem of free will.