Gottfried Seebass: The Significance of Free Will

-- The Determinism and Freedom Philosophy Website --

Prof. Dr. Seebass, of the Philosophy Department of the University of Konstanz, begins by putting aside the discussion of determinism and freedom in terms of the doctrines of compatibilism and incompatibilism. He supposes, no doubt rightly, if for his own reasons, that we need to start again. This he does by beginning with what it is to will something -- to have a conscious optative attitude. The freedom of our willing is a matter of two dimensions, one of these being what is natural or essential to a person. This naturalness is the central concern of the paper. Certainly the paper is outside of the circle of stuff that is readily understood by students of philosophy in the English langage. That is not a reason for not working at it. What is needed with determinism and freedom is an escape from the declining industry of compatibilism, incompatibilism and the compromises and variants. Here is one.


The Significance of  “Freewill“ is one of the great, old and possibly “eternal“ enigmas of mankind. In 1437 the Renaissance philosopher Lorenzo Valla diagnosed that there is “scarcely a question more in need of being answered and less close to an answer at the same time”. Since then, I suspect, things have not changed essentially.

Yet many people think otherwise. Some are unconvinced of the topic's oldness. Some scholars have claimed that “freewill” is an invention of Judaeo-Christian theology unknown to the Greek tradition. This claim, however, proves false on both counts. Others believe it to have been solved long ago or antiquated. Schlick in 1930 and Schopenhauer almost a century earlier expressed anger at all those “ignorants” who “still spill much ink” on a subject they believed had been long since settled. And more recently Davidson dismissed it bluntly by citing a number of philosophical predecessors who had done “what can be done, or ought ever to have been needed, to remove the confusions that can make determinism seem to frustrate freedom”.

Now, the history of philosophy is full of examples of alleged “definite solutions” that turned out to be over-hasty. “Freewill” is a case in point. The widespread tendency to dismiss it stems from a double error. On the one hand, its significance tends to be underrated due for the most part to a misconception of, or resigned disengagement from, what it really means to be active. On the other hand it is badly overrated in that it is assumed, explicitly or implicitly, that “willing freely” means ipso facto “being entirely indetermined in what one wills” or (in the words of Kant) being independent “of the mechanism of nature in its entirety”.

The latter misconception is mirrored vividly by the established division into “compatibilism” and “incompatibilism”. This distinction suggests that it is one and the same thesis that is affirmed by the one party and denied by the other. But this is misleading. First, the question of determination is only one of a number of different aspects of “willing freely” and is not in itself a question of “all or nothing”. Second and most important, what is at stake among the two parties is not whether one should accept or reject the thesis that “freedom and determinism are compatible”. The real issue is how the concept of “freedom” relevant to understanding ourselves as active beings, or as beings that are responsible for what they do, should be analyzed or explicated. Consequently, the thesis affirmed by the so-called “compatibilists” is not the same thesis that is denied by the “incompatibilists”. Also what kind of determination the latter want to deny and for what reasons is an open question allowing for various answers.

So let us forget from now on this unhappy categorization and go to work afresh. As elsewhere in philosophy, this work is mainly conceptual. We need to know what we are talking about in discussing “freedom of will”. Of course we do not want to know what philosopher X or scientist Y have said these words mean. There would be no end to listing different definitions and declarations. We want to know what we ordinarily mean when talking of the “free will” of active and responsible persons. And to know this we have to clarify the constitutive notions, beginning with “will”, “volition” and “willing”.


Let us state first what willing is not. Obviously it is not a brain state or brain process. I know that I want to finish my talk, and you may want to raise some objections in the discussion. But neither I nor you know what corresponds to this in our brains. That there is something going on up there, most probably something very complex, is not in doubt. But for the purpose of identifying our volitions these facts are irrelevant.

Still, to acknowledge this conceptual truism does not prejudge the questions of further explanation and possible ontological reduction. Couldn’t it be the case that willing to finish a talk is wholly dependent on, or even identical with, a complex brain pattern? Maybe, but to show this two essential steps must be taken. First we would need to establish general, stable and specific correlations between conscious volitions and brain events, hopefully one-to-one correlations or else one-to-many, yielding some kind of “supervenience”. Some people believe we are pretty close to this already. Now, I am not a neurologist and do not know the vast amount of relevant empirical findings. However, due to my earlier work on language and thought and my recent cooperation with neuro-psychologists I am not totally unexperienced in reading neurological studies and have looked at some of them more closely. So I have every reason to be sceptical about enthusiastic, far-reaching claims. The permanent, all over activity and interconnectedness of the brain make it difficult to single out and interpret relevant patterns even if these are localized fairly well. So it is unlikely from the start that specific neurological correlates of intentional states like “wanting” or “willing that p” can be identified by the present experimental methods. And to my knowledge there is no concrete proposal up to now.

This holds true also and most specifically of the readiness-potential measured by Libet and his followers. Setting aside the well-known conceptual and methodological difficulties of these experiments there are two principal obstacles. The findings are much too unspecific to represent concrete volitional states. Moreover, they cannot be the relevant neurological correlates, if (as is reported) they are prior to conscious volition. At the very most they can be causes of other (unknown) brain patterns that really correspond to intentional will. So we would not be better off anyway. The only rational stance towards ongoing neurological research on human volition is: wait and see!

Even more problematic and futuristic is the second step. Given all the correlations we need, how are they to be interpreted? If the relation is irreducibly that of event causality involving a time-difference, dualism is inevitable, although some philosophers will insist (denying the necessity of some “loss of energy”) that the efficiency here runs only from the physical to the mental. Shunning dualism other philosophers claim that what we have here is identity: token identity at least or even type identity after the manner of chemical reductions like “water is H2O” (or some variant weakened by “supervenience”). Although advocates of these positions assure us that they know what they are talking about, I doubt it. The idea is strange indeed. How physical objects, like neurons or an entire neuronal network, might be the direct bearers not only of relevant physical properties, but also of mental properties like “willing” or “wanting that p”, is an ontological mystery that needs mystics to be believed. I am unsuited for this task. To my mind, the only ontologically respectable alternative to dualism is neutral monism as advocated by Spinoza, Fechner and Mach. This is a metaphysical position, too. But it is consistent and no conceptual mongrel.

The main problem for the volitional “naturalist” is epiphenomenalism. This covers more than the classical causal accounts given by Huxley and Nietzsche. Mental events are in danger of becoming epiphenomenal, too, if they are taken to be mere correlates (as in classical psycho-physical parallelism) or secondary qualitive aspects conjoined factually (for whatever reasons) to something accounted for completely in physical terms. Thus epiphenomenalism is likely to follow from neutral monism (recognized firmly by Mach) and from all sorts of alleged “identity theories”.

This is a threat, of course. For the idea that our conscious volitions, beliefs and deliberations are just accidental performances on an “inner stage” caused by, or conjoined to, hidden neural processes getting along on their own is the epitome of irrationalism. Nietzsche and his postmodern adherents are quite happy with this. Others shamefully try to overlook this result or to avoid it by inconsistence. But irrationalism is the price one has to pay for unbounded physiological “naturalism”. If this were our natural human condition, the idea of being active and responsible would be an accident of our “inner theatre”, too, as Nietzsche and his followers have claimed all along. Fortunately, all of this is “science or philosophical fiction” up to now, despite all pseudo-scientific rumour. Lacking further evidence it is wise to keep one's head and to remember (varying Hamlet) that ”there are more things in heaven and earth than are dreamt of in your physiology”.


If willing is not a brain state or process what is it? Behaviorists such as Hull or Berlyne proposed a dispositional analysis of intentional states like believing or willing, referring to interconnected internal states that are ascribed to human beings as “hypothetical constructs” (in the sense of Meehl and MacCorquodale), viz. states or events thought to be realized physiologically. In a modified form this conception still survives under the name of “functionalism”. Now, though it is true that willing, in contrast to idle wishing, is qualified dispositionally, the idea of its complete dispositional explication is hopeless. My detailed reasons for this are given in a book to which I must refer those of you who are inclined to think that this still is a plausible option. Surely willing, believing and other intentional states can not be defined by the various “causal roles” they play in connecting relevant inner states or mediating between external stimuli and responses.

The core of willing, stripped of its motivational and possible further qualifications, is a conscious optative attitude related reciprocally to an assertoric attitude as its conceptual counterpart. Both states are elucidated in a preliminary way by the metaphor of “directions of fit” (in the vein of Frege, Anscombe, Searle, and Tugendhat). Much more can be said, however, about what this means and in what ways wanting, willing, deciding or intending differ from the mere optative. Still, as is the case with every basic notion, there are irreducible elements in it that cannot be analyzed further, but must be grasped by the individual learning the relevant terms of language. Talk of unconscious or preconscious volition is derivative from its conscious form. Thus in the discussion of “freewill” we are referred to the conscious, intentional and motivationally qualified state of willing.


Turning next to the concept of “freedom” I will take a shortcut. In the most general sense, being “free” means being “unhindered”. Hobbes and Schopenhauer have argued for this, and their diagnosis is confirmed by ordinary language. More specific senses of “freedom” are subsumed easily. This refers in particular to the classical notions of “freedom of action” and “freedom from constraint” developed by Greek philosophy. Most forms of social, political and economic freedom are covered by these. Freedom of action means: being in a position to act as one wants or wills to act. Freedom from constraint is more complex, meaning in a broad sense almost the same as being unhindered but having several narrower meanings that overlap partly (already hinted at by Aristotle) with what we would call “freedom of will”. Freedom of will is another special case branching again into various subcases. First, however, let us look more closely at the concept of “hindrance” itself.

Clearly this is very general notion to be specified best by the questions what is hindered in what and by what. A river that is not hindered in its flow by dams or embankments is said to “flow freely”. A paralyzed or tied man is “unfree” because he is hindered by abnormal (internal or external) impediments to move as he wants. The hindrance need not extend to every part or aspect and need not be absolute. Still it must be significant. Common to all subcases is the idea that something is restrained severely from evolving, living or existing in ways which are “natural” or “essential” to it and therefore should not be precluded. Hence whether, in what respects and to what degree something is free or unfree has to be judged from two dimensions: the relevant standard of “naturalness” and the relevant realm of theoretical possibilities actually closed or open. Roughly, the greater the number of possibilities open, the lesser the hindrance and the greater the freedom. However, possibilities that do not touch on the relevant “nature” can be ruled out as inessential. Most of the infinite number of actions I might envisage but actually cannot do are irrelevant to my freedom as I do not and will never care about them in the least. In the limiting case even a maximum of open possibilities can mean no freedom at all if they are totally “inessential”. Conversely, freedom may not be diminished if the only possibility open is just the only “essential” one.


Concerning freedom of action it is tempting to think that the question of “naturalness” or “essentialness” is settled by what one wills. Let us be content with this for the moment and see what determinism would mean for “acting freely”.

As has been worked out lucidly by Aristotle, determinism is not confined to causal or nomological types. Rather, determinism is the thesis that every event is fixed without alternative (“determinatum ad unum”) and therefore necessary in some sense, depending on the specific reasons for this fixation. More precisely, for every pair of (elementary or complex) propositions “p”/“¬p” and corresponding states of affairs, the realm of possibilities open is reduced to one, the relevant alternatives being excluded definitely. Determinism is an ontological thesis, not an epistemic one and thus independent of predictability. Also it is a general thesis covering mental as well as physical events. This was recognized by the early Stoics and Epicureans. Later on Christian theology came to realize that universal determinism is implied by the belief in an omniscient and almighty god. Nevertheless some theologians thought, as did the later Stoics, that volition could be exempted. Both types of thought are still virulent today.


Suppose first that mental events can be undetermined while the physical world is fixed. Then in what sense can we say that “unhindered” physical actions are “free”? Clearly, the possibilities criterion fails. Measured by this standard freedom has a zero degree. So we are left with the second criterion, “naturalness” or “essentialness”. Assuming that volition defines what is “essential” we could say that we are free if the will and the physical world (say: the movements of our lips in speaking or the motionlessness of our arms when voting) coincide. Otherwise we are unfree. If volition is unfixed and flexible we may increase freedom by adjusting it to the world. This is what the later Stoics and their followers recommend. The advice is not easily complied with in some cases (e.g. speaking and voting). But anyhow: Isn’t a reactively adapted will unfree, too, as its formation is constrained by physical reality? And doesn’t the fact that we will have to do what is fixed anyway, independent of our will, rule out freedom of action at least in the physical world, even if our volition is undetermined and merely happens to conform to what is fixed?

Most people will accept these conclusions. Some, however, try to evade them. Thus theologians such as Origen and Molina contrived an argument, revived more recently by Harry Frankfurt, which is intended to show that the inexistence of open alternatives is not detrimental to freedom. The argument is flawed, though, and not difficult to refute. It seems convincing because a volition free of adaptive constraint and coinciding with what is fixed seems to be realized without hindrance. Recall, however, that this depends on the premiss that this will is the only “essential” one at the relevant instant. How is this to be safeguarded? Well, people believing in a benevolent fate or a god who knows best and takes care of their “real persons” at every instant are on the safe side. But do we believe in this or some secular substitute of it? If a fire breaks out and I find the doors locked, would I really think that my freely and successfully not having wanted to check them earlier is part of my “nature” and that my present constrained will to escape is but a fleeting misunderstanding giving way readily (fate permitting) to my considered insight that it is “essential” to me to be burned within a few minutes? Certainly not.

Accordingly, Locke stated long ago, following common opinion, that a man locked in a room is not free to leave it even if he does not care about leaving at present. Being able to leave or stay according as to what one wants surely is “natural” for a human being. And the deeper reason for Locke’s important point had already been uncovered by Plato, Aristotle, Cicero, Plotinus and Augustine. Modified slightly and reformulated in my terms it is this: Even if it is not denied that human volition is a necessary, or even decisive, criterion of “naturalness”, one cannot confine it to actual states but must take into account others that could be formed under certain conditions, viz. conditions that cannot be ruled out as “inessential”.

The lesson is clear. Leaving aside strong metaphysical presuppositions (or very special cases where there really is only a single volitional state that proves to be “essential”), we would not ordinarily call any action “free” that is voluntary but does not depend on the will. And if we would, or rather would have it that way for philosophical reasons, it is quite clear that this notion of “freedom” is not relevant to the claim that we are active beings and responsible for the outcome of our reactions.


Now let us drop any reservations concerning the mental and turn to an uncompromising, universal determinism in the vein of the early Stoics and the mainstream of Christian theology. According to such a conception, all of what we think, believe and will, all of our practical and theoretical reasonings are fixed without alternative. Whether this is due to the order of fate or god, to natural laws, or due simply to ontological reduction is immaterial. All that counts is the fact of fixation and its implication for freedom of will and action.

Now, it is certainly a mistake to define “freewill” in an indeterminist man­ner. The adequate definition is: “freedom to form one's will”. This covers many different processes most of which include deterministic parts and elements, though they are not normally thought to be fixed throughout. Forming the will may consist in a spontaneous single act but need not. It may imply extended weighing of reasons and calculating means, ends and consequences. But it may also be confined to a quick decision. A more extensively reflecting person may try to stir up (time permitting) unconscious attitudes in the endeavour to find out what she “really wants” or “wants inmost” or to arrive at some “volitional equilibrium”. Also reflection may or may not extend to volitions of a higher order, as has been countenanced by Augustine, Abelard and more recently Frankfurt. All of these processes may be carried out under restraint, external as well as internal, and with a significant lack of knowledge. If so, they are considered traditionally (beginning with Plato and Aristotle) as being hindered and unfree since the resulting will is precluded more or less from relevant alternatives. Otherwise they are presumed to be free (lacking further hindrances).

Suppose, however, that all alternatives are ruled out and the entire process is fixed. Could we still call it “free”? Obviously there are problems here (noticed at the latest by Aristotle, Epicurus and the Stoics). Accepting determinism for causal and theological reasons Augustine attempted to solve them. To save the ordinary assumption that freedom of action implies alternative options he invented the conditional analysis of the concept of practical possibility which allegedly ensures this formally. And to ensure the freedom of the conditioning will, he contrived a formal argument to the effect that volition is free ipso facto. Both of these theoretical devices have been adopted by myriads of theologians and philosophers up to now, at least in part and in a modified form. Ignorant of their theological origin many of the so-called “compatibilists” rely on them in the belief that they are achievements of modern philosophers such as Hobbes, Hume and Moore.

However, despite the respectable authority of St. Augustine and all of his followers, both moves are flawed. The argument for intrinsic volitional freedom is invalid. And the conditional analysis is no analysis of practical possibility at all as it does not extend to its relevant modal sense and merely shifts problems back to volition, albeit perhaps to volitions of a higher order. The crucial questions are still the same: Having dismissed the possibilities criterion radically, we are left with the criterion of the “naturalness” or “essentialness” of volition as a last resort. But, leaving aside Augustine’s argument or metaphysical premises, how could this be provided?


Basically two solutions have been suggested. First it is said that the will is free if its formation satisfies certain general patterns which can be regarded as “natural”. Some philosophers have been content even with the unspecific condition of being “amenable to reasons”. Others, along with jurists, have tried to specify lists of characteristic epistemic defects and restraints the absence of which defines, allegedly, what “willing freely” means. This may be tolerable for practical purposes, say in the courts. But it is surely inadequate theoretically. On the one hand, it is indifferent to the optative attitudes entering into an informed and unrestrained formation process. Also it is confined to the average, ignoring what is “essential” for a particular person at a particular time. On the other hand, granting the definition as a mere stipulation, we still get no answer to the most central point: Why should we say of someone determined (by god, fate or natural causes) to satisfy the relevant patterns that he is active and responsible in this case any more than in another case where he happens to be determined not to fulfil them? Clearly, this notion of freedom, contrived for practical reasons at best, is squarely beside the point.


The second solution, proposed independently or in support of the first, looks even more Augustinian: Couldn’t we simply say that volitional states and processes are “essential” for some person ipso facto if they are taken to be defining? This would be a nice solution indeed. However, what volitions will qualify? Arguing in a sweeping manner (which actually is but a secular variant of the belief in a universally caring god) Kant and Schopenhauer proposed that the entire set of our volitions and actions defines the “empirical character”, indicating successively (to ourselves and to others as well) who we are as a person. But apart from the transcendental metaphysics backing this generous form of “personal essentialism”, the idea is strange. Obviously the multifarious trivialities and contingencies of our lives are not constitutive for us as persons. Also we do not register passively only what volitions might crop up in our minds.

Yet the same passivity is characteristic of the idea that there might be some stable, limited set of attitudes to be discovered (by reflection or progressive self-experience) to be constitutive of what we are. This form of “personal essentialism” is not as strange as the other one. But it is still peculiar. If it were true, we should be able to list at least some of the most basic attitudes supposed to be definitional. But are you able to do this? And on what grounds could the defining attitudes be selected? Merely because of their factual (genetic, educational or habitual) entrenchment? Hardly. Such attitudes can be taken as well, and are taken frequently, to be mere conditions under which we live or even hindrances to our living and acting as we would want. To make them part of our personality, we have to accept them, i.e. decisively to “appropriate” them or to “identify ourselves” with them.

Hobbes thought that preserving biological life is the only unchanging object of our guiding will. Bentham, Freud and many others thought that it is gain and preservation of pleasure. But a short look at human history makes clear that even these very general and unspecific goals, unfit to single out individuals anyway, do not qualify as defining personal characteristics. An Al Quaida suicide terrorist surely does not care very much for his physical life and pleasure. Personal life is an ongoing dynamic process made up not only of passive experiences and receptions but also of instances of actively forming our attitudes towards facts and possibilities. This is the central element in setting the standards of what is “essential” for us and therewith of relevant hindrances. Accordingly, they are subject to change. A volition we firmly “identify with” at present may well be dissociated later on. And even if some basic volitions were to be stable and definitive of what we are, it would certainly be absurd to claim that all of our concrete willings and doings are free and unhindered only to the extent they can safely be referred to those. All of this concurs to the conclusion that the second solution, when considered more closely, proves to be badly inadequate, too.


Suppose, however, the ordinary conviction that we can have an active part in the development of volition is false (say: for epiphenomenalistic reasons) no less than is, supposedly, our common belief in the existence of objective alternatives. If so, the ordinary notion of “freedom” would not apply to us, provided that what is hindered is still an individual person and not some volitional or other state declared independent of her activity (e.g. by a better-knowing god) to be “essential”. Moreover, if what we want and do, as well as the ways by which we come to it, are fixed in advance and in every detail, we cannot be responsible actors. And as this condition is implied by the traditional system of normative social control, different from merely manipulatory forms of shaping behaviour, this system will become inapplicable, too.

In fact the consequences are even more radical. They lead to a fatalist view of life. Of course this does not refer to the so-called fatalism of “lazy reason”, better called “foolish reason”, as it rests on the denial of obvious causal dependencies. This idea is ill-conceived and not consequently deterministic. Fatalism, rightly conceived, includes the mental realm and amounts to the thesis that the course of the world, being fixed throughout, cannot be altered (theoretically) and therefore cannot be influenced actively (in the relevant practical sense of “can”). Within philosophy this should have been settled ever since Aristotle. More recently Richard Taylor has renewed the same argument provoking dozens of articles which in vain have tried to refute it. Clearly, a consistent determinist must be a fatalist − provided he is determined happily to be consistent.


In view of this, philosophers having a temper similar to that of Schlick and Schopenhauer might grow angry about the fact that there are still so many “ignorants” “spilling ink” in the hopeless endeavour to get around what is obvious by contriving, following the lead of St. Augustine, some substitute notion of “freedom” that fits a fixed universe and still preserves understanding ourselves as active and responsible persons.

In fact, these hopeless endeavours are motivated by a double belief: (1) that determinism is true, or might well be true, as the universe is confined to what can be accounted for in strictly “naturalistic” terms and therefore must be seen as “causally” or “nomologically closed”, and (2) that the only way to deny determination is to affirm “blind chance”. Now, I am not in a position to prove or refute either determinism or naturalism. But I think that it is possible to deny both of these and that to do so is (at the very least) not as desperate as are the continuing illusions or self-deceptions about the consequences that would result from their truth for rationality, activity, responsibility, and personal freedom. Whether microphysical indeterminism, taken to be implied (pace David Bohm) by standard quantum theory, could be of any use to freewill is unclear to me. Some of the relevant proposals (e.g. that of Kane or Hameroff/Penrose) are certainly subject to the objection of “blind chance”. Others might well be better off. But I know of none which is satisfying. So I must leave it with this. Before ending my talk however, I would like to draw your attention to one further important point.

Even confining ourselves to the physical world and natural science, it is amazing to see how many people still believe that determinism, though unproven, must be true. It is not just the existence of quantum physics. Any experimental physicist knows the insuperable dispersions in measurement. Most theories rely on probabilistic rather, than deterministic laws, if they formulate general laws at all and do not confine themselves to regularities and correlations. All of this holds all the more strongly if we move up to psychological, social and cultural phenomena. Mankind has always lived in a world full of probabilities and contingencies not amenable to deterministic calculation. To believe that this is merely the result of epistemic defects that could be overcome is (at the very best) a useful methodological maxim and apart from that, as a general ontological thesis, simply an ideology.

To me the only understandable reason for this is fear of “blind chance”. Surely our world is not full of this, though chance, apparently, cannot be excluded. But anyhow, to deny that something is determined is not in the least equivalent to affirming that it happens “by mere chance”. The false belief to the contrary is one of the big blunders made repeatedly ever since the early critics of Epicurus. The existence of statistically constant marriage and suicide rates, e.g., proves neither that individual marriages and suicides are matters of mere chance, nor that some people are determined somehow to marry or to kill themselves (as believed by some of the early 19th century interpreters of “moral statistics”). Both kinds of events may well result from undetermined free decisions of individual persons. Indeterministic freedom may not exist in reality, but it is surely not inconceivable or conceivable only at the expense of being “chancy”. This argument against those who are still interested in making sense of understanding ourselves as free, active and responsible persons, this bad old ploy at least should disappear completely from the discussion.

HOME to Det & Free website front page
HOME to T.H. website front page