Freedom Beyond Metaphysics

About: Olivier Boulnois, Généalogie de la liberté, Seuil

Olivier Boulnois’ latest book, Généalogie de la liberté (Geneaology of freedom), is a history of the concept of freedom from Aristotle through Kant to the present that seeks to answer two vast and fundamental questions: Are we free? And how should freedom be understood? As Boulnois himself admits in the introduction to the book, this task is as ambitious as it is unachievable. He nevertheless approaches this history by adopting a well-defined perspective: The antinomy between freedom and determinism is inextricable because it is a false problem born of a forgetting—the forgetting of freedom itself. Through this forgetting, the counterpart of Heidegger’s forgetting of being, a primarily ethical notion was gradually transformed into a metaphysical concept, namely, the concept of free will, understood as the power to choose an action or its opposite. According to Boulnois, the original philosophical sin is the idea that moral responsibility requires a metaphysical foundation in free will, which is thus reduced to the role of prerequisite or condition for the possibility of responsibility. From this perspective, history is that of a progressive veiling of the “authentic” sense of freedom initially identified and enunciated by Aristotle. Thus, the book traces the main stages of this veiling since the emergence of the Aristotelian conception of freedom. After Alexander of Aphrodisias and Augustine established the concept of free will in late antiquity, medieval thinkers of the thirteenth and fourteenth centuries, from Thomas Aquinas to John Duns Scotus and from Henry of Ghent to William of Ockham through Peter of John Olivi, provided lasting structure to the problem of freedom by means of a series of oppositions: nature versus will, intellect versus will, divine predestination versus human freedom. These oppositions structured the field of reflection on freedom in the modern era, from Descartes to the third antinomy of the Critique of Pure Reason and, beyond Kant, to debates in analytic philosophy that followed the fault lines clearly defined by the divide between compatibilism and incompatibilism—a divide that covered, respectively, the “libertarianism” of the supporters of free will and the determinism of its opponents. According to the author, these fault lines are not self-evident, but are derived from the conceptual structure inherited from this long history.

Thus, Boulnois undertakes an essentially negative work that paves the way for a positive, “ethical” conception of freedom according to which true freedom is a form of happiness. This conception, which is merely sketched out in the book, entails above all the capacity of an agent or subject to adhere to the good, with ethics guiding his or her action and thus making it free (p. 478).

It is impossible here to review the entire argument, as it covers an immense historical corpus that ranges from Aristotle to Leibniz and engages with later thinkers such as Kant, Nietzsche, Wittgenstein, and Chisholm. I shall therefore only touch on some highlights of the analysis. Incidentally, the regrettable and surprising absence of a bibliography or index makes it difficult to navigate this scholarly work, although the detailed table of contents and the (largely chronological) plan partly compensate for this shortcoming.

Responsibility Without Free Will?

According to Boulnois, the dilemma between free will and determinism, of which Kant’s third antinomy is the best-known formulation, rests on two premises: 1) the confusion of causes and reasons for action, namely, the idea that the reasons for a choice are the causes of the action that results from said choice; 2) the incompatibility between being caused and being free, which imposes a choice between the affirmation of freedom and the principle of causality, and even the principle of reason. These two premises are not justified in the eyes of the author—who, in line with Wittgenstein and Anscombe, invites us instead “to escape the causal mechanism and the concept of subjectivity in order to think freedom” (p. 84). It follows from this that the problem of free will is a false problem whose solution amounts in fact to a dissolution.

This is where the analysis gives way to genealogy proper. Dissolving this false problem requires showing that such a framing of the question of freedom, far from being self-evident, is the result of a history that did not begin with the idea of free will. Thus, Boulnois defends the thesis that Aristotle put forth a complete and coherent conception of action that did not involve any notion of free will. He also argues that a correct understanding of this conception would bring us closer to the “authentic phenomenon of freedom” (p. 106). Indeed, rather than opposing nature to the will, Aristotle saw voluntary action as a particular form of animal movement. Actions that can be attributed to us are those performed, not “in spite of ourselves,” but “voluntarily” (hekousion), which is to say, without constraint or ignorance. Such actions depend on us (eph’ hemin) and presuppose a prior prohairesis, which the author understands as a “decision.” Although prohairesis is the principle of action, it is not a first cause and it is dependent on principles. To be sure, Aristotle maintained that we are responsible for our actions, provided that we have the possibility of not performing them. Yet, as Boulnois observes, the rational power of opposites is merely generic: The physician has the capacity to heal or to harm in general, but not the capacity to accomplish the opposite of what he is doing in specific circumstances (for instance, harming a patient if he has judged it appropriate to heal him). Aristotle’s theory does not require a central decision-making faculty (the will), nor does it need to reconcile the existence of such a faculty with physical determinism. Aristotelian freedom, which is conceived on the model of the free man in the city, is not a metaphysical attribute; rather, it is part of an ethical horizon and it consists in the ability to do good.

The Invention of Free Will

So why and how did the concepts of the will and free will appear? Their emergence is the result of a three-stage process: First, the Stoics’ invention of the will as a decision-making faculty (to eph’ hemin), albeit within a deterministic framework; second, Alexander of Aphrodisias’s invention of free will; third, Augustine’s attribution of free will to the will. In the wake of Suzanne Bobzien and Michael Frede, [1] Boulnois argues that Alexander was both the key thinker in the genealogy of free will and the first libertarian philosopher. Contra the Stoics, he invented the concept of free will through opposing fate and that which depends on us, namely, that which results from a free choice between contingent alternatives. Human beings are free because they enjoy the power of opposites, a power that is not only generic (the power to heal or to harm), but singular (depending on circumstances, representations, and desires, the physician may wish to heal or to harm). The free agent has the capacity to perform the opposite action, and responsibility presupposes that the agent could (or, in some cases, should) have chosen otherwise. Thus, Alexander invented free will as a prerequisite of responsibility while also separating freedom from its ethical dimension.

What is Augustine’s contribution to this history? Above all, he made freedom an attribute of the will. For Boulnois, however, the interest of Augustine’s thought lies in a well-known aporia: On the one hand, the will is absolutely free and is able to will both good and evil; on the other, the will does not have the capacity to will the good on its own unless God has determined it to do so—as Augustine never ceased to emphasize. This tension paved the way for two opposing and incompatible interpretations: For some, Augustine was a libertarian; for others, he was a compatibilist and even a determinist who prefigured Luther’s concept of the bondage of the will. According to Boulnois, the tension inherent in Augustine’s thought is a striking illustration of the problem posed by the articulation of freedom with ethics. The concept of free will has a “fundamentally dimorphic character,” for it must reconcile the idea of an indeterminate power of opposites with an essential orientation towards the good.

The “Long Middle Ages” and the Clarification of the Question of Free Will

Although medieval authors did not introduce any radical changes, the grafting of Augustinian ideas onto the Aristotelian heritage caused them to reflect on freedom through the lens of its relation to natural causality, determination by desire, and divine predestination. However, not every theoretical position could be defended in these debates, and no medieval thinker would have endorsed determinism. Deterministic positions could indeed only emerge through theoretical reconstruction: Not only were they never affirmed, they were continuously attacked by their opponents, as illustrated by the condemnation of 1277 at the University of Paris. Among the 219 theses denounced as contrary to the faith by a commission presided over by the Bishop of Paris, Étienne Tempier, a number concerned free will—for instance, proposition 131, which stated that in the presence of a desirable object the will is obliged to will it. Some medieval historiographers identified Siger of Brabant (circa 1240-1284) as the author of the condemned proposition. In reality, Siger considered that a cause can always be prevented from producing its effect. He also maintained that we are free because reason, which enables deliberation, gives us the power to prevent the action of a cause. Accordingly, he did not deny the existence of free will, nor did he defend a deterministic conception of causality.

The question of the compatibility of free will with natural causes was indeed addressed in medieval debates; however, it was overshadowed by two other questions. The first concerned whether free will lies in the intellect or in the will. Siger and the early Thomas Aquinas stressed the primary role of the intellect, whereas Henry of Ghent and John Duns Scotus emphasized that of the will. Boulnois describes how the opposition between intellectualism and voluntarism emerged in the context of post-Aquinas debates. He also examines the case of the Franciscan Peter of John Olivi (circa 1248-1298) who, more than anyone else, valued the will as the only active power and who affirmed that man is rational because he is free.

The second question concerned the relation to divine predestination. Latin medieval thinkers agreed that the human will has causal power over its actions. The key issue, then, was to reconcile the causality of the human creature with the causality of God. Aquinas did so by claiming that God acts (“moves”) on the created will, but according to His own nature, without this will being necessary. The majority of later thinkers rejected this solution as incoherent. Instead, like Olivi, they asserted that the will, as a free power enjoying the freedom of indifference, could not be determined by anything, not even by God.

The Many Paths to Freedom and the Limits of Grand Narratives

Behind the complexity of material and conceptual transmissions and the vast array of problems tackled, the genealogy of freedom proposed by Boulnois tells a rather simple story: that of the progressive veiling of a conception of responsibility, considered in its purely ethical dimension, under a metaphysical faculty, the will, characterized by absolute freedom. The villains in this narrative are the libertarians, with Boulnois implicitly embracing a position quite similar to that of Peter Strawson: The libertarians make responsibility hostage, as it were, to a metaphysical and scientific question—“Is the world deterministic?”—when in fact it is a fundamentally ethical concept.
While the libertarian stance is entirely defensible on a conceptual level, the author diagnoses it as a theoretical impasse. This diagnosis, however, is open to challenge. Indeed, over the last fifty years or so (since the publication of Harry Frankfurt’s famous article “Alternate Possibilities and Moral Responsibility”), the focus has shifted to examining conditions of responsibility that are independent of the capacity to act otherwise, or, following Strawson, to analyzing the attitudes and practices that underpin our attributions of responsibility. [2] In addition, freedom is no longer only conceived as a negative metaphysical condition (as the absence of determination) of moral responsibility. Many philosophers in the analytic field (again, starting with Frankfurt in his seminal article “Freedom of the Will and the Concept of a Person”) have engaged in positive reflection on freedom, understood as self-determination or as rational self-government. It would have been interesting to compare these conceptions of freedom as autonomy with the historical conceptions analyzed in the book, all of which link freedom to adherence to the good, the latter being, for Boulnois, the last word of freedom—perfect freedom, understood as liberation from all bondage. Indeed, contrary to the thesis put forward in the book’s conclusion, the concept of autonomy does not immediately imply adherence to an objective and substantial conception of the good. [3]

A second remark concerns the historical scenario presented in the book: Free will was born in late Antiquity, the various aspects of the antinomy between determinism and freedom were definitively established during the Middle Ages, and these continued to structure debates up to contemporary analytical philosophy. It is undeniable that many of the same aspects played out in the Middle Ages and the modern era. But does this mean that the moderns contributed nothing, or that they merely reenacted the same aporia as the earlier philosophers? The controversy between Hobbes and Bramhall over freedom and necessity is virtually absent from the scenario recounted in the book. And yet, this controversy played a major historical role. In some respects, it mirrored medieval debates: It focused on the compatibility of free will with antecedent causes and with divine foreknowledge and providence. However, it also shifted and deeply transformed the terms of the debates. [4] First of all, Hobbes and Bramhall had radically opposed conceptions of agency. Bramhall claimed that the rationality of man made it impossible to reduce human action to animal movement. Drawing on Duns Scotus, Aquinas, Aristotle, and even Plato, he saw the will as a rational appetite guided by objective norms and values. Hobbes, on the other hand, refused to distinguish between human action and animal behavior: He argued that both are subject to the same mechanisms of determination of action by beliefs and desires. Thus, the controversy between Bramhall and Hobbes not only concerned the relationship between free will and determinism, but also revolved around the question of naturalism. Bramhall sought to preserve a fundamental gap between the sphere of action, which is subject to reasons, and that of natural causality. Hobbes rejected this duality and conceived of action in terms of mechanical causality. This first divergence was due to a radical opposition in the very conception of causality. For the Aristotelian Bramhall, a cause is an agent endowed with capacities or powers through which it acts. This theory of causality allowed him to reconcile the conception of free will as the power of opposites with the principle of causality. On the contrary, for Hobbes, a sufficient cause is a set of prerequisites that imply the existence of an effect. An agent is a body which as such is not a cause; it is so only insofar as its movement is determined.

Accordingly, Hobbes rejected all conceptions of agent causation and replaced them with a conception of event causation. Clearly, then, the dispute between Hobbes and Bramhall was not a mere repetition of medieval debates, nor did it simply reenact the controversies born of the Reformation: It revealed the crucial role played by the conception of causality and the adherence to naturalism, a fault line that was at least as fundamental as determinism.

Several other points would merit detailed discussion. For instance, one might wonder whether it is truly possible to save the baby (moral responsibility) while throwing out the bathwater (free will), and whether it would not be preferable to take seriously, in the manner of Spinoza, the idea that moral responsibility and its associated moral conceptions presuppose free will and must be abandoned, or at least profoundly revised, if free will is an illusion. One of the main strengths of Généalogie de la liberté is that it gives rise to this sort of question, confirming that it is not just a book on the history of philosophy, but also a work of philosophy in and of itself.