Keywords

1 Delineating Agency

‘Agency’ is a word with multiple meanings. As it certainly does not ambiguously denote a psychological category, I begin with clarifications.

The word ‘agent’ in English has been in use since the seventeenth century to identify a factor or power held to cause a change. For ‘agency’, the Oxford dictionary cites Darwin, who wrote about the pollination of flowers ‘requiring the agency of certain insects to bring pollen from one flower to another’. ‘Agency’ denotes capacity and power attributed to matter (as in chemical ‘reagent’), to institutions or social organizations with the power to act on behalf of people (like ‘the Central Intelligence Agency’), and to people individually (as in ‘the agent of her own destruction’). It has also been common to refer to God’s agency or the agency of spirits. Historically, the opposite of ‘agent’ was ‘patient’, and this is a reminder of the active/passive distinction of the premodern ontology to which the language of agency is heir.

The multiplicity of references to agents of different kinds persists in everyday speech about causation. The language of agency permits descriptions of naturalistic and non-naturalistic, material and mental, individual and social causes to exist alongside and in interaction with each other. In everyday usage, it is possible to refer to everything from car brakes to rental company to driver error to intention to kill as the principal agent of an accident. However, there has been a historical shift, so that there is a modern emphasis on using the word ‘agency’ to denote a moral or political, that is to say, distinctively human category. As a result, speakers now might reserve the attribution of agency to people or institutions in a car accident (to continue with the example) and distinguish other factors as contributing causes. This modern usage describes the agents as individual people or groups of people who are said to have the power to be the cause of events, and said to have the power of self-direction . Agency has become linked to notions of the autonomous self and to the dignity or status accorded to a ‘free agent’. Thus, it is a notion important to moral and political issues. Feminists, for example, place great weight on women acquiring agency and in critiquing the circumstances in which this is constrained. Indeed, much of the experienced meaning of agency derives from its opposition to the notion of constraint. Beyond this, the notion of agency also has an influential place in humanist and existential philosophy, with expression in psychology, where reference to agency denotes something like a reference to freedom as a defining condition of being human.

It is important to be aware of the range of linguistic usage. It is possible to refer both to material or spiritual things as agents and to people, distinctively, as agents. Before the twentieth century, language sometimes described people as agents, but in doing so it ordinarily attributed the agency to the will, to the soul, or to reason, rather than to a psychological subject or a self. In a parallel way, some recent writers attribute agency to the body, to the unconscious, or to the brain. All the same, contemporary writers commonly attribute agency to people, or to the self (as in the statement, ‘I did this’), and they attach a special value to it. These multiple usages are rather confusing for analysis, if unexceptional in everyday speech.

In the light of these comments, I stress one point. Reference to agency in twentieth or twenty-first century psychology, as in the human sciences generally, may simultaneously invoke what are generally thought of as causal processes and what are thought of as free actions. At first glance, as a result, it would seem as if psychologists are deeply equivocal about accepting or denying free will . I suggest that actually there is no deep-lying confusion behind this equivocation, if that it be; rather, there is something special in the projects of psychology, namely, their ability to provide description and analysis appropriate for understanding people, as opposed to understanding brains (in terms of deterministic causal processes) on the one hand, and juridical, moral, or political subjects (with imputed absolute freedom of action) on the other hand.

The modern notion of the individual person as agent first developed in legal, political and theological contexts. The history is intimately connected to the developing notion of a self. Reference to human agency denoted, and still denotes, action originated by individual legal, moral, and political subjects, or by institutions viewed as analogous to individual subjects, acting without special or noteworthy constraint. This has often been called free agency. The Oxford dictionary, in this context, cites Coleridge’s political demand that ‘the State shall leave the largest portion of personal free agency to each of its citizens, that is compatible with the free agency of all’. It is the normative practices of politics and morality that have made it important to distinguish human agents from other agents. Insofar as psychologists have been drawn into discussions of human agency, they have taken part in these normative practices.

In consequence, it makes no sense to ask for a non-evaluative account of a person’s agency. Discussion of agency, in the last analysis, involves questions about the relation of human subjects, intentions, and evaluations, and of the language and culture which are their expression, to the causal material world that is the subject of the natural sciences. This is why agency is a problematic category for natural scientists who think that science excludes evaluative judgments.

When psychologists refer to agency, they use a psychosocial category. Just as there is no non-evaluative use of the word ‘agency’, the word has no psychological meaning independent of social content. The literature about agency, as a psychosocial category, therefore inevitably takes positions on the long-standing question of the relation between psychological and social forms of explanation and the institutionalization of those forms in separate psychological science and social sciences. The discipline of social psychology clearly faces these matters most directly (and there has even been a psychological social psychology and a sociological social psychology; Good 2000).Footnote 1

While ordinary people, and often enough psychologists too, might now say that a person is an agent (and explain the agency, say, by reference to intelligence or cognitive capacities), the social scientist understands agency as attributed to a person (or institutions or things). In the language of social science, agency is a status not a state (Barnes 2000).Footnote 2 In a related way, in the language of many analytic philosophers, will, agency, and choice are not powers but human actions under certain kinds of descriptions. From the viewpoint of these disciplines, it is a misusage, a category mistake, to talk as if agency were a psychological state or psychological power. Rather, it is a power of persons (the classic statement is Ryle 1949/1963). Applying this lesson in psychology, we can say that when psychologists talk about agency, they utilize a category with social content and take part in the process of social and political ascription of status to people. But it muddling matter. This is because reference to agency as a status ascription (or attribution) persists alongside and interacts with the older usage in which reference to an agent denoted a capacity or power (whether material, mental or spiritual). Moreover, there are, of course, psychologists with religious beliefs or who uphold a humanist philosophical anthropology, for whom agency is indeed a ‘real’ state, a state valued and thought essential to being fully human.

There is nothing contradictory, then, though it may be confusing, to describe the body as the agent of a person’s desires, while at the same time describing something in the body (an illness perhaps) as constraining a person’s agency. By virtue of its flexibility and open-endedness, everyday language about agency has a rich instrumentality.Footnote 3 For example, it makes sense to talk about training or disciplining the body in order to build up patterns of behaviour (or habits) understood in causal terms, in order to give a person more agency, understood as freedom of action. Contemporary advocates of the brain sciences promise that new knowledge (of causal processes) will give individuals more agency (freedom of action; Rose 2007, 2013). Indeed, this is the common pattern of argument of the psychological and social sciences over the past couple of centuries: Let us understand human nature (causally) in order to improve human wellbeing (free agency). All the same, there have been and there are religious and humanist critics of this, the enlightenment project, who hold, for instance, that the very act of investing in causal explanatory language about people derogates from the morality and politics of free human agency .

To the extent that agency, understood as a category that is properly applied to persons, is now a flourishing interest for psychologists, this, surely, is yet another manifestation of the individualism characteristic of Western modernity. Ways of life informed by psychology are bound up with concern for the power and capacity individuals have and do not have in the social worlds they inhabit. It might be thought that as researchers in a field of natural science, psychologists could and should substitute the notion of cause for the notion of agency in describing and explaining behaviour. As a family of fields of research about people, and even more as a family of practices concerned with everyday individual capacity and activity, however, psychology has a large place for the category of agency. Even if psychologists have at times carried on as if the human subject were not inherently social (as if brains existed in vats, for example, not in social people), this is not possible for those who adopt the language of agency. It is most obvious, perhaps, in psychotherapy—in all its multifarious forms. Therapeutic reference to agency invokes knowledge of the ‘internal’ powers and constraints of a person, along with the ‘external’ powers and constraints of the social world the person inhabits.

Notions of agency and constraint are of manifest significance in everyday psychology, as a conversation taking place between two of the people involved with the volume in which this chapter appears illustrates. (Whether my account is accurate is not important now, and I have used my imagination.) Their interest was whether lack of constraint on individual actions in at least some areas of Western society, allowing individuals the freedom ‘to do their own thing’, or, as the advert says, ‘just do it’, might not actually be a constraint. Intense individualism is indeed not easy for many people. The discussion was not about either autonomous will or neurons but enlarged on a psychological, everyday approach to the psychosocial conditions of people’s lives. It was talk about people, not wills or brains, as agents. Such talk has a history and a social specificity .

The complexity of language about agency is compounded when there is an implied reference (as in the citation from Coleridge) to free action. Indeed, I rather think connotations of free action always colour the word ‘agency’. I do not propose to tie myself or my readers up in the free will/determinism debate . But, I will note how psychological as well as everyday language about agency often combines description of a person as the site of power bringing about an effect (analogous to saying water is the agent of dissolution of a salt) and the description of a person as a free agent (for example, a person responsible in law). The language of agency does not dictate a position on the free will/determinism question: it holds things open and allows further evidence and judgment to refine description and explanation. Everyday usage allows for fudge, continuous negotiation, infinite shades, in attributing agency understood as both cause in the scientific sense and as the kind of free power people attribute to people able to carry out legal and moral acts. There are, to be sure, radical libertarians who emphasise the extent of agency understood as individual free power. And there are severely reductionist scientists, by contrast, for whom there can be no such thing as free agency. Many people in Western cultures, I presume, take a position somewhere between these extremes, and in doing so, they think of human agency as the activity of a person (relatively) without hindrance or constraint from other powers, a person who has experience of variable degrees of actual agency. In this middle position, it is common for ‘agency’ to mean something like ‘the real cause’ or ‘the principal cause’ of something happening, that is, the cause which on a particular social occasion is thought to be the key to the actual unfolding of events. Thus, someone might refer to the brakes not the driver as the agent, the real cause, of an accident, though of course both brakes and driver belong in a full list of causes. (The full list, literally speaking, requires the history of the world.) Philosophers discuss what I am calling the broad middle position under the heading of ‘compatibilism’, the label for the view that it is not inconsistent to accept determinist causal explanations in the natural (or social) sciences while upholding belief about the free action of people (Kane 2005).

I am suggesting that there are constructive psychological ways of referring to the agency of individuals, individuals with the power or capacity thought intrinsic to being human in general and to being the persons they are in particular. This psychological usage contrasts with polarized alternatives. These alternatives represent the two poles of Cartesian dualism: attributing agency to mechanistic body, and hence attributing all change to the physical agency of matter in motion; or attributing agency to a soul imbued with transcendent powers. Psychological reference to agency, at its best, displaces this polarity. Ordinary speech in psychological society, and much psychological discourse both lay and professional, refers actions to people not to the operation of mechanistic processes or souls. It uses language incompatible, for instance, with the belief, upheld in some versions of popular neuroscience, that there is no human agency or personal volition because ‘we are our brains’. But psychological reference to agency does not presuppose that agency is the expression of a specific power of the soul or faculty of mind, like the will, understood as somehow acting independently of physical processes. The psychological approach to agency is historically an outgrowth from, and continues to blend with, everyday understandings of the powers that people have and the constraints under which these powers operate. These understandings have multiple roots in political, religious, ethical, legal and philosophical culture. Only in recent centuries, perhaps we might say in the past two centuries, have these roots nourished a distinctively psychological representation of the agency.

Modern studies of agency as an attribution recognize its intrinsically social content, and, with that, recognize the intimate connection of attribution of agency and ascription of responsibility. In many contexts, certainly in judicial settings, to attribute agency is to ascribe responsibility. As the British sociologist Barry Barnes argues: ‘An understanding of the everyday employment of this concept [of responsibility], with its double significance —psychologically understood it implies internal capacities, sociologically understood it implies liability and answerability—is…the key to an understanding of the role of ‘choice’, ‘agency’ and related concepts in everyday contexts’ (Barnes 2000, p. 1). Consider also the word ‘Victorian’. It is a word now coloured by knowledge of ardent nineteenth century convictions about moral agency, which was thought to be exercised through individual strength of will, strength of character, and by the fact that the presence or absence of will was thought to validate judgments about the responsibility of individuals . It is worth bearing in mind the insightful, if exaggerated, comment of one late Victorian writer, who said that it is only the social question of responsibility and punishment that sustains public interest in the question of free will (Hyslop 1894, pp. 181–182; Smith 2013a, p. 165). Very often, we will find that when the talk is about agency, the talk is at base about responsibility.

2 Volition and Psychological Agency

There are many ways of conceptualizing and representing agency. Earlier ages treated spiritual powers, the devil, the stars, and the passions as agents. There are, of course, modern communities that continue to make such attributions. The history of specifically psychological conceptions of agency encompasses the field of volition , desire and will—the division of conation, alongside cognition and affection, in traditional tripartite description of mind. In this section, I contribute a historical sketch.

There is a curious interest in this for the modern psychologist because volition largely disappeared from view as a topic for professional psychologists during much of the twentieth century. Since the 1970s, there has been some revival (as called for in Kimble & Perlmuter 1970). Experimental psychologists transferred an interest in agency, understood as the motor of behaviour, from mental will to drive, personality, cognitive processes, the emotions, the unconscious, or, alternatively, to the body—and in recent decades to the brain. Large areas of psychology went ahead without concern for or even with active antagonism to the notion of volition .Footnote 4 Yet at the same time, therapeutic, educational, organizational, and counselling psychological practices all deployed some notion of psychological agency, even of volition, even if implicitly rather than in an openly systematic and theorized form. Notions of volitional agency flourished even more in the domain of so-called popular psychology, in the world of know yourself and self-help books . In the contemporary world, and above all in the world of imputed consumer choice (‘just do it’) there is a strong focus on agency, agency which in ordinary speech often enough appears as an expression of will. The word ‘agency’ is indeed widely used to denote individual power in actual or desired forms of economic exchange and governance. Most significantly, in respect of social power, legal systems continue to pose decisions about responsibility in terms which people understand as referring to internal mental states of volitional agency. Presumably, the renewed interest among some psychologists in volition as a category reflects all this.

The early intellectual history centres on belief in the intrinsically active power of the soul to cause actions. Christians in the early church sought knowledge of the will in the light of their understanding of the relationship between the will of God (with omnipotent agency) and the action of the human soul (owing to sin, with restricted agency) . The question of freedom of the will, that is, the extent of human agency independent of God’s grace, was a source of deep divisions from the time of the later Stoics and of Augustine (Frede 2011). It came to a head again during the Reformation, and it was the subject of a famous exchange between Erasmus and Luther.Footnote 5 There was a large Renaissance literature on more secular views of volition, particularly as applied to the control of the passions. Shakespeare’s plays commented on the role of fate or fortunà, as opposed to the will, in a person’s life. Belief in the power or agency of the individual soul was a source of dignity and, for some, the very basis of civilized, as opposed to barbarian, existence. In his discussion of the passions, Descartes wrote: ‘I see only one thing in us which could give us good reason for esteeming ourselves, namely, the exercise of our free will and the control which we have over our volitions . For we can reasonably be praised or blamed only for actions that depend on this free will’ (Descartes 1649/1985, p. 384).Footnote 6 There was an association between the will, agency, dignity and emerging notions of the identity and worth of the individual self.Footnote 7

The actual word ‘agent’ spread with the expansion and increased precision of natural philosophy in the seventeenth century. The word denoted powers or capacities without drawing sharp distinctions between spiritual, mental or material agency. Old theological disputes and new science together ensured the continuation of debate about the place of freedom and ‘necessity’ in the course of things. The word ‘determinism’ , applied to nature and to human affairs, came into use only after about 1870 (Hacking 1983).

Locke’s prime concern in his Essay Concerning Human Understanding (1690) was the sources and certainty of knowledge. His book nevertheless included a very influential statement on the origin (or agent) of human action, which he attributed to pleasure and pain. Empiricist writers subsequently pictured individual agency as the response of a person to the pleasures and pains of experience.Footnote 8 Whether, and in what sense, this implicated necessity was much debated. Some authors, like Joseph Priestley in England and C.-A. Helvétius in France, embraced necessity and, indeed, they built their hopes for enhanced individual political and moral agency, and thus for enlightened political and social life on this basis. Writing in the same spirit, Jeremy Bentham established political utilitarianism, a scheme for a legal system which would appropriately distribute pleasure and pain, and hence regulate individual agency and thereby ensure a rational ordering of human affairs. He expected people to be rational and hence to choose his system: he assumed that natural human rationality conferred agency. The many opponents of utilitarianism argued, however, that belief in necessity was destructive of morality and incompatible with common-sense awareness of each individual’s power to will actions. They also doubted the extent to which rationality in fact conferred moral or political agency. The more conservative of these critics turned to the moral will, informed by religious faith and enforced by royal and religious power, as the source of agency. Indeed, in the eighteenth and nineteenth centuries, there was an evangelically informed turn in Britain and in the USA to the authority of subjective knowledge of agency, understood as a mental power, ‘the will’. Most authors in English, at least before the second third of the nineteenth century, did not conceive of these discussions as being about psychological agency in any distinguishable sense; rather, their subject was human nature in all its social, political, psychological, economic, legal, medical, linguistic, and philosophical dimensions. Agency was distributed as a subject in what was called moral philosophy (and, later, the moral sciences), and there was no specific discussion of agency as a category in its own right.

A profound and, in the long run, influential expression of belief in the intrinsic agentive character of the mental world, which later commentators did not hesitate to call a contribution to psychology, appeared in the writings of Maine de Biran in the first decade of the nineteenth century. Biran judged awareness of effort to be the source of the most elementary and irreducible knowledge of mind, and he therefore rendered mental agency—‘l’effort voulu’—constitutive of the personal mind or self.Footnote 9 He published piecemeal and he left his essays unfinished; nevertheless, later writers understood his work to originate a distinctively French psychological–philosophical view of agency, developed most prominently in Bergson’s and in Sartre’s (completely different) accounts of free agency as defining what it was to be human.

The work of Victor Cousin was actually more influential in France in the first half of the nineteenth century. It acquired formal standing in the French higher education system as a result of Cousin’s position at the head of the institutional structure for teaching teachers. Making a broad claim for ‘psychologie’ as the road to philosophy, Cousin, like Biran, stressed the irreducible volitional character of the self; but whereas Biran used language expressive of an almost phenomenological awareness or subjectivity, Cousin used a more recognizably traditional Christian language referring to the power of the soul. When Taine and Ribot turned against this psychology in the 1870s, promoting what Ribot called ‘the new psychology’ (an eclectic mixture of German experimental psychology, physiological psychology, clinical evidence, and British associationism), the stage was then set, as elsewhere at the time, for debate about the implications of developments in mental science for agency and determinism .Footnote 10

According to a least one historian of philosophy, Kant bequeathed the antimony of freedom and necessity as ‘the great problem of modern thought’ (Pinkard 2002, p. 43). When he discussed the nature of human beings, Kant separated ‘the human’ as an anthropological subject, for which empirical argument was appropriate, and discussion of the essentially human as the agent of the moral law, for which he turned to transcendental reason. His anthropological writings and lectures (which subsumed psychology) discussed human nature and activity in a manner that flowed together with everyday talk in the educated circles in which he and his students moved. He included discussion of capacity and agency in connection with habit, mental disorders, constraint, character, and so forth. His formal moral discourse, by contrast, defined the condition of being human in terms of an abstract, absolute imperative, and in this context, freedom denoted the human obligation and power to use reason and act according to the moral law. This propounded a moral theory of obligation and not a psychological theory of capacity or agency. All the same, Kant’s arguments underwrote and legitimated an understanding of agency as a capacity of the human spirit, the position that informed German-language philosophical anthropology in the nineteenth century and beyond. The claim that agency, embedded in the human spirit and expressed through language and culture, was constitutive of being human lasted till, but hardly survived, the horrors of the twentieth century. It certainly fostered psychological formulations, as one can see, for example, in C. G. Jung’s project in scientific psychology (in which Jung sought to juxtapose his own conclusions about the collective unconscious with Kant’s thought; Shamdasani 2003).

When John Stuart Mill published A System of Logic, the book that provided the philosophical underpinnings for utilitarian social and political thought, he dealt directly with the issue of necessity. He took the position that was to be the mainstay of modern philosophical argument (the position, as noted above, called compatibilism). In Mill’s presentation of the case, a person’s actions were said to be free when they were the actions of the whole person, unconstrained in any significant way. Certainly, Mill held, pain or pleasure was the proximate cause of a movement or of behaviour, but this fact was compatible with saying that a person had agency (or, as the Victorians would have said, showed character and exercised will) if the movement or conduct expressed the purposes and character of the whole person (Mill 1843/1900, Book VI, Chap. 2). A person in prison did not have agency in the way a person at home did, even if both responded to pleasures and pains. Mill was a political libertarian and a moralist committed to creating conditions that would enhance personal agency, but he believed in both necessity and personal freedom while strongly opposing any notion of the will as some kind of spiritual force.

During Mill’s lifetime, that is, through the middle years of the nineteenth century, a large specialist and popular medical and scientific literature spread the conviction that a person’s agency might not be nearly as extensive as those who stressed the role of the will, Mill included, tended to assume. There was considerable interest in instincts, habits, and automatism, in the hypnotic state and in spiritualism, and in disorders implying a loss of control from drunkenness to epilepsy. A physiological psychology developed, and its promoters built on the model of reflex action to propose new scientific understanding in bodily terms of much that had earlier been thought attributable to the activity of mind. In this connection, the London physiologist W. B. Carpenter, in 1853, introduced the idea of what he called ideomotor action, to describe the way an idea in the mind, or an obsessive thought, caused activity over which a person had no control and of which a person may even have been unconscious (Carpenter 1853, p. 672). In effect, Carpenter (and other medical writers) redescribed the mental act of anticipation leading to a movement as a physical (higher brain) reflex. The physiological language merged with a large popular literature, replete with vivid case studies, that discussed when and where individual agency, understood to depend on will, as an empirical matter of fact did or did not exist (Smith 2013a). There was an English language automatism debate, prompted by T. H. Huxley’s lecture on the topic in 1874, and this was the intellectual background of William James’s intense personal and theoretical psychological interest in the will (Huxley 1874/1894). James took over Carpenter’s language of ideomotor action and argued that action follows directly from the presence of an idea in the mind and does not require a separate mental effort or will force (James 1890/1950, vol. 2, pp. 522–528). He set this psychophysiological discussion of will in a larger framework of argument, however, and this larger framework advanced into another version of compatibilism. While stressing the physical determinants of psychological life (as he did in his theory of the emotions as well as of the will), James upheld ‘the will to believe’ as a moral project, indeed as a project without which he as a person could not live well, and this emphatically defended a notion of individual agency.

Though it was a cliché of the time to describe the nineteenth century as an age of science, belief in free will did not just persist in this age but even gained in strength from what was thought to be empirical knowledge. A large part of the popular appeal of phrenology in the 1820s and 1830s, for instance, was that people believed knowledge of cerebral capacities as the determinants of character and gave individuals power to strengthen or control mental life for themselves (Wyhe 2004). Though accused of materialism and hence immorality, phrenology increased not decreased a sense of personal agency for those who thought it true. Analogously, in the course of the evolutionary debates, leading advocates of scientific naturalism like Huxley and Francis Galton promoted belief in the uniformity of causation in nature and human nature as part of a moral crusade to empower individuals and society alike on the basis of ‘the facts’. It was the message of Huxley’s much commented on lecture, ‘Evolution and Ethics’, which ended by quoting lines from the late English national poet, Tennyson: ‘We are grown men, and must play the man “strong in will/To strive, to seek, to find, and not to yield”’ (Huxley 1893/1989, p. 86). Science supported a moral project of enhancing agency. Galton’s studies of the contribution of heredity to character, which had a determinist form, were matched by a political programme, eugenics, to do something about it. Medical demonstrations of automatism in illnesses like epilepsy, in hypnotic trance, or in habitual drunkenness, fed into Victorian rhetoric about the necessity of will power—the cultivation of personal agency.

There were numerous philosophically oriented attempts to overcome what seemed to many people to be confusions of thought in this mixture of scientific and moral culture. This debate about the nature and possibility of free will in a scientific age was important to the way psychology developed as a field in Britain, Europe, and the USA . It sustained links between psychology and mental philosophy just when some psychologists were looking towards experimentation as the means to make psychology a distinct scientific field. An integrated commitment to scientific knowledge and willed agency was a striking feature of the late Victorian age.

If we turn to the everyday language of human relations, the language, for example, of the nineteenth-century novel, we find a mixture of the attribution to causes (habit, social conditions, sex, heredity, and so on) and attribution to individual will to be not just ordinary but ubiquitous. From the 1830s or so, English-language writers increasingly, but never exclusively, articulated what they had to say about such topics under the heading of psychology. Victorian authors continuously mixed and negotiated description of mental forces, such as will power, and causal determinants, such as training, age, social circumstance, illness, and custom. Stirring the mixture often enough supplied the novelist with a plot. It was a mixture of seemingly infinite variety. While holding that young children were not agents and hence not legally responsible, both Catholics and Evangelical Protestants stressed the presence of the will in children and held it to be an innate moral or spiritual force, a force for both good (as self-help texts presupposed) and bad (as references to wilfulness and to upbringing designed to break the child’s will made clear).Footnote 11 A large moralistic literature about individual agency attended to the practical powers that individuals might possess, acquire, or lose. Strictly consistent explanation in the language of mental forces or in the language of causal bodily processes was not a priority. Discussion of agency was marked by flexibility of description and openness to negotiation according to particular circumstances that observers thought prevailed. There were, to be sure, times where the flexibility broke down, and I discuss in the next section the criminal court where a defence of insanity sometimes congealed and polarized opinion.

There was a noteworthy struggle over the representation of agency in Tsarist Russia, which influenced the way psychology developed in that country (Sirotkina & Smith 2012). The rigidly autocratic political system upheld faith in the soul as the bearer of personal agency and ground for attributing individual responsibility . Social order in the autocracy appeared to require this faith. During the brief period of relative liberalization under Alexander II, in 1863, I. M. Sechenov published the first version of his article on ‘The Reflexes of the Brain’ and N. G. Chernyshevsky published a notorious novel, What Is To Be Done? Sechenov, a physiologist trained in German and Austrian laboratories, turned to the model of reflex action in order to imagine a physiological analogue for the mental process of volition. When criticized for removing the grounds for belief in individual responsibility, he denied that this was the implication of his argument. For Sechenov, knowledge of the natural bodily conditions of life was the basis for the exercise of agency, not faith in the soul. He did not use a political language of agency or even a moral language of free will ; indeed, he was not permitted to do so under conditions of censorship.Footnote 12 All the same, he contributed to a political programme to replace the theocratic agency of the one tsar with the enlightened rational agency of many informed subjects. Even though his language described nervous and psychological processes, readers understood the message. Chernyshevsky’s novel was more direct and, remarkably, still published (Chernyschevsky 1863/1989). His story brought to life a group of young people who had adopted a full-fledged rational egoism, the position that for Chernyshevsky expressed reasoned agency. He attributed to reason the agency to act on behalf of the natural needs of the person, to unite body and mind, and to escape the fetters of irrational passion and ignorant and repressive moral and religious codes. Liberated individual actions informed by reason, he maintained, would ensure collective progress.

These writings were part of a public debate, in the course of which they faced two of the most widely read and influential ripostes to such notions of agency in the nineteenth century. Interestingly, in everyday language now it would be a commonplace to say that these ripostes showed much greater psychological insight, even if their authors did not in any way describe themselves as psychologists or contribute to academic psychology. Turgenev, in his novel Fathers and Children, the publication of which preceded Sechenov’s and Chernyshevsky’s contributions, portrayed a student of medical physiology, Bazarov, who unsentimentally dismissed the possibility of any kind of agency that did not have the form of a physical agent (Turgenev 1862/1960). His own agency as a proponent of this materialist worldview, however, was tragically cut short when he fell in love and when he contracted an infection that killed him. His own conception of agency was pathetically inadequate for his own life. Independently, Dostoevsky, enraged by Chernyshevsky’s novel, in a response poorly understood at the time because it appeared so antihumanist, wrote Letters from the Underworld. The Letters purported to be written by a man (significantly with no name) who, in opposition to any conceivable constraint, and most painfully in opposition to the gift to him of actual, profound love, asserted the arbitrariness of his own will (Dostoevsky 1865/1960; Frank 1986, Chap. 21). Dostoevsky’s antihero was an agent, come what may. The Letters thus bitterly parodied rational egoism—they pictured the will as self-destructive as well as destructive of others. Dostoevsky’s view of agency was radically anti-enlightenment and pictured agency as a source of tragedy in the human condition.

Within the family of Aleksandr I. Herzen, the Russian political exile, there was actually a living and not fictional debate between father and son. The father, who belonged to the romantic generation coming to maturity in the 1830s, faced by necessitarian physiological argument, turned to human history and social progress to legitimate a conception of agency. His son, Alexandre A. Herzen, flushed with enthusiasm for the physiological science of the 1850s (and a future professor of physiology in Lausanne), upheld determinism (Sirotkina 2002).Footnote 13 Neither was a psychologist in any specialist sense; but such debate about agency over the years played a large part in shaping psychology as a public field.

When experimental psychology began to acquire a place in German universities, followed by other countries, its academic proponents inherited the antimony of freedom and necessity. Wundt, to take a key instance, maintained a separation between causal analysis thought appropriate at the level of psychophysiological phenomena, and a form of understanding in terms of mental apperceptive and conative activity though appropriate at the level of higher mental events (Danziger 2001). He elaborated a psychology with, in effect, a place for active, agentive mind or spirit, and this psychology existed alongside the contributions he made to physiological psychology. He developed research in the former under the heading of Völkerpsychologie, a science that turned to language, myth, and cultural life generally as the collective expression of human agency . In other hands, especially at Würzburg under the leadership of Külpe, by contrast, in the years before 1914, there was research to make volition a rigorously examined experimental topic, though one could hardly describe this as in any direct way about agency. Albert Michotte also studied in detail the immediate antecedents of voluntary choice (studies which preceded his better known work on the perception of causation; Michotte & Prüm 1911). With the establishment of a psychology discipline in universities and colleges in the USA, which involved considerable concern for the scientific standing of the field, volition gradually became sidelined as a topic. North-American psychologists turned to the motivation, intelligence or personality, rather than volitional agency, of their subjects, to represent activity in ways judged to be scientific.

For the German idealists, Fichte, Schelling, Hegel, and (for these purposes) Schopenhauer, and in the books of many more accessible exponents of idealist worldviews, like E. von Hartmann, agency was not a specifically psychological topic. Rather, idealists sought to characterize the dynamic form of existence as such and only secondarily to explain the agency of a particular person as an individual. Idealist thought, however, underwrote belief that there is agency, though this was variously understood to have the form of impersonal powers (perhaps unconscious forces, perhaps a ‘will to power’), or to have the form of a psychological expression of spirit in the individual mind. Awareness of the potential conflict between agency understood as reason and agency understood as non-rational force contributed to the literature of intellectual crisis , which Nietzsche did so much to deepen, that was so prominent around 1900.

Elements of the idealist intellectual tradition persisted and influenced humanistic forms of psychology in the post-1945 period. Existentialist thought stressed ontological freedom, the ultimate agentive character of the human condition and this was taken up in psychological terms, for example, in the work of Erich Fromm. Informed by Christian ethics rather than European philosophy, Rogerian therapy, built on the client-centred principle of nondirective regard, was a practical enactment of the agentive ontological status attributed to clients.

What was experienced as an intellectual crisis , because the foundations of reason themselves appeared to be questioned, was the context of new thought, in great variety, about the unconscious. Freudian psychoanalysis strongly emphasized the agency of the unconscious, rather than the agency of rational conscious capacities, in action and character formation. In contrast to earlier notions of unconscious events, which pointed to the role of unconscious anticipation, the new psychologists of the unconscious pointed to the burden of repressed memory (individual and inherited; Hayward 2014). Freud and his followers, indeed, portrayed this as a new step in human self-understanding . It is therefore very striking that Freud, who was committed to a strong psychological determinism (complementing physiological determinism), also remained committed, like a good Kantian philosopher, to the power of reason to stand apart from unconscious forces, to comprehend them, and thereby to offer at least some hope for human freedom (Tauber 2010, 2013). Freud acquired a reputation as the hammer of bourgeois confidence in the power of the will, since people are, he claimed, driven by unconscious forces; yet, at the same time, he reasserted a Kantian view of agency as the possession of reason to discern the moral law. In this way, I would argue, he exemplified enlightenment thought as it developed in the psychological and social sciences. In these sciences, a notion of agency persisted, the agency of the reasoning subject given institutional form in the world of science, even while the activity of scientific reason was creating knowledge exposing the causal determination underlying what people did. The resulting dilemmas have been acknowledged and explored in political thought. For example, Sonia Kruks has persuasively discussed Simone de Beauvoir’s work as a long engagement with the political ambiguity of the human subject understood as both causally situated and agentive, both conditioned and free (Kruks 2012, Chap. 1). De Beauvoir’s art in exploring this as an irreducible ambiguity, finding a literary as well as philosophical voice, brought what she had to say close to the practical understanding of people in everyday relations. This ambiguity persists, I would affirm, in the modernist project, which asserts the causal necessity of events alongside the demand that individuals live one way rather than another.

The philosophical conundrum here has given rise to a huge literature, much of it refining different stances in the position known as compatibilism.Footnote 14 Ordinary people continue by and large to be compatibilists, as Barnes notes: ‘Much of our everyday discourse manifests a robust compatibilism, in that it is content to regard actions as at once chosen and caused’ (Barnes 2000, p. 4).

Modern psychological discussion of free agency, in line with compatibilist argument, has taken the form of empirical studies of everyday ways of ordering the world in terms of perceived causes. Social psychology research in the second half of the twentieth century shifted towards the study of the cognitive and emotional processes involved in people’s judgment and understanding of the agentive character of what people did. This research therefore did not directly address the question—which came to appear ontological rather than scientific in nature—as to whether people were agents. Debate was not about what agency ‘was’ or to what extent people were ‘really’ free agents, but about how people viewed agency in terms of personal traits (including their own) varying between individuals and even varying within one person in different circumstances and over time. Attribution theorists, for example, in the 1970s and 1980s, researched what people perceived and said in everyday life in order to understand causal cognition and attendant moral judgment (Jaspars, Fincham, and Hewstone 1983; Hewstone 1989).Footnote 15 It would seem that this brought social psychology somewhat closer to a social science approach, in which agency was understood as a status attribution. There was, besides, interest in finding psychological tools to increase personal agency. Beyond this, there was discussion of the personal and social advantages of belief in agency.Footnote 16

The dismissal by twentieth-century psychologists of volition as a category was part of a general suspicion about explaining human capacities by internal mental states or processes. Many sociologists, often influenced by Wittgenstein, shared this suspicion and understood references to mental states as a certain kind of language game or social activity. The sceptical psychologists supposed that human activity was at base the activity of physical systems: there was no mental power, certainly no free will .Footnote 17 The sceptical sociologists argued that references to mental states were status attributions, descriptions of social relations in particular ways of life. From the sociological point of view, agency was a collective achievement expressed in social order built on the regulatory notion of individual responsibility . For large numbers of psychologists, agency was a function of physiological capacities. For one large group, the social psychologists whom I have just mentioned, research was interested in what people believed rather than in agency as a ‘real’ state. Yet, in spite of all this and whatever the power of the arguments, everyday psychological discourse, and the discourse of a good deal of expert psychology along with it (therapy, guidance, forms of training, and such like) continued in fact to refer, routinely, to internal states and, if not so much specifically to will or volition, to desire, intention, motive, purpose, and choice.

Both history and ethnology provide a comparative perspective on when and how psychological practices advance or damage agency—and of which kinds of agency and for whom.Footnote 18 The work of the sociologist Nikolas Rose is an influential reference point in discussion of liberal democracies (Rose 1985, 1989/1999, 1996/1998). Under the rubric of ‘the history of the present’, Rose described psychological practices as distinctively modern forms of governance in societies of a kind in which many people, but very far from all, have come to live, that located the power or agency maintaining social order within individuals and within occupations (psychotherapy, educational guidance, counselling, etc.) that worked to ensure people were responsible individuals. During the course of the twentieth century, the Victorian emphasis on will and will power was reshaped as advice and training in the techniques of self-management . When such psychological practices were at work in governance on a broad front, we can identify psychological society, a society in which there is a strong tendency to assign agency to individuals understood as psychological beings (Smith 2013b, Chap. 4). Rose was substantially concerned with the question where agency was ascribed—conscious that this linked the history of psychology to politics. His arguments, sociological in form, located agency not in individual selves or minds but in the practices themselves. ‘To account for the capacity to act one needs no theory of the subject prior to and resistant to that which would capture it—such capacities for action emerge out of the specific regimes and technologies that machinate humans in diverse ways’ (Rose 1996/1998, pp. 186–187).Footnote 19 There have been humanist critics who saw such analysis (which was linked both by Rose and by his critics with the work of Michel Foucault) as damaging notions of individual agency; but Rose argued that agency lay with the enhancement and availability of practices not with some imagined ‘internal’ reality (Rose 2007; Derkson 2011).Footnote 20

The history of psychology is also a source of studies of the agency of individual subjects of psychological practices—the patients in Freud’s case studies, for instance, or the participants in psychological experiments, some famous, like ‘Little Albert’ (Harris 1979, 2011) or like the students in Milgram’s research on obedience.Footnote 21 These studies open up the politically and ethically significant topic of how in fact psychologists have themselves ascribed agency to the people with whom they interact, with what consequences, and what co-operation or opposition from participating subjects (or ‘patients’). This contributes to a critical, politically and ethically reflexive, psychology.

3 The Insanity Defence: Debate on Agency Exemplified?

In this section, I try briefly to provide a more precise historical case study of what debate about agency may mean in practice. Exemplary demonstration of a number of the points made in this essay is to be found, perhaps, in an area where statements with psychological (or psychiatric) content interact with legal decision making. I say, ‘perhaps’, because the question of agency in connection with the plea of insanity in criminal cases (the focus of the debate here) has attracted a huge amount of comment, and comment has returned again and again to the issues as if the ground had not been gone over many times before. Clearly, the issues continued to be troubling .

Western codified and common law criminal legal systems maintain that the establishment of guilt requires a demonstration both that a certain thing happened (actus reus) and that the accused party was an actor in the events with a certain capacity of mind (mens rea).Footnote 22 In everyday language, we might say that a finding of guilt requires a defendant to have been the agent of the relevant event.Footnote 23 As a result, over the centuries, there has been debate about what makes a person a legal agent, and statements have centred on mental states and capacities. Legal writers have construed these states and capacities variously, sometimes as formal legal categories with no specific empirical referent, sometimes as empirically verifiable states of mind, intentionality , or social attribution. They have understood the discussion to be a matter for jurisprudence, not at base, a matter for psychology.

When physiological approaches to mind became common in the nineteenth century, creating psychological medicine, proponents sometimes said that science had disproved the existence of free will (in individual cases or even in general) and, as a result, the administration of justice and punishment had to change. The same argument is occasionally heard now in the wake of the huge growth of research in the neurosciences. The argument was, and is, vitiated by at least two misapprehensions. First, statements that causes determine effects are not empirical conclusions of natural science but logical expressions of the form of knowledge established in natural science. All events, scientifically understood, have causes: finding the causes is the name of the game. This holds just as much for sociological or psychological as for physiological research. Second, it follows, free will is not a meaningful notion in natural science discourse (or in social science or psychological discourse understood as a form of natural science) but, rather, belongs in philosophical, theological, moral, political, jurisprudential, and everyday psychological discourse (along with a good deal of applied psychology discourse as well). In these latter discourses, it is persons, not brains, or bodies, or even minds, to whom free agency is ascribed. In sum, new knowledge about the brain, however extensive, leaves the question of free will or free agency where it was before, since the brain is not the kind of thing that can be said to have free agency.

With the legal setting in mind, I think we can conclude that the large issue for debate is not free will versus determinism but the relationship between natural science knowledge (including psychological knowledge that has the form of natural science knowledge) and the kind of discourse that politics and jurisprudence—and everyday psychology and a lot of professional psychology too—articulate. Whether someone has agency in a political or juridical sense is a matter for debate within the languages and social practices of politics and jurisprudence. Of course, there has been and is a place in this for medicine, for psychology, and for social science. The role of these disciplines is to provide clear and accurate descriptions, in terms that doctors, psychologists, and social scientists agree (or try to agree) are the best within their respective fields, of the capacities and illnesses individual people do or do not have. Provided with such descriptions, courts, administrators, and ordinary people are then in the best position to judge whether or not a person (or indeed an institution) has the status that legal rules and custom identify as agency. The decision is not a matter of empirical proof, however much the rhetoric of justification may use empirical language, but of formal or informal rule following by using empirical evidence. It is not the presence of illness in itself that leads to an acquittal in a criminal trial but the convention that a legal system may allow evidence of psychotic illness to signify that a person does not have legal capacity (Morse 2004, 2007).

It is illuminating to draw a parallel with the legal standing of children. The age of a child is part of what makes a child the kind of person who can be said to have agency of the legally appropriate kind, or not. The law has to draw distinctions between ages when a child is and is not responsible in relevant ways. Different jurisdictions have different rules, and this is a matter of social morality and custom not of natural science knowledge. Everyday responses to what children do, operate with highly flexible and negotiable language about the relative agency of children. It is not the age of the child that makes the child free or not free but, rather, the social process of attributing free agency, a social process that takes the growing child to have changing degrees of agency.

A parallel kind of flexibility, allowing for at least some element of negotiation, has become characteristic of the legal administration of mental illness in criminal courts in the past century and a half or so. There are, for example, procedures for pre-trial determination of mental illness and unfitness to plead, there is in England, since 1957 (following Scottish example), the much used plea of diminished responsibility, and there are possibilities for transferring defendants after trial between prison and hospital. All these add up, in principle, to a flexible social means for arbitrating the status of agent and of patient. (How flexible and just it is in practice is not now the point of argument.)

Full of enthusiasm for the advance of physiology, some Victorian medical witnesses in the courtroom made the naïve point that scientific progress, since it had made it possible to describe causal events in the body, enabled experts to say whether a defendant was constrained by illness to do what he or she did and hence to say whether he or she was responsible. Defence lawyers played a substantial part in trying to put such argument before juries (as they have done more recently, for example, in connection with compulsive eating disorders; Eigen 1995). Victorian doctors, however, as they gained experience of giving evidence about insanity and faced hostility when they claimed that their knowledge proved legal incapacity, came to understand the legal context in which they performed. On the occasions when Victorian doctors, citing progress in physiology, claimed that the courts should take much more notice of or even defer to scientific evidence about insanity, judges made two persuasive objections. The first was that if insanity were explicable in causal terms and as a result of this, insane actions were not culpable, all human conduct would be similarly explicable and not culpable. The idea that no action was culpable was dismissed. Indeed, few people have ever wanted to promote it; the Lombrosian school of criminology is a major exception.Footnote 24 The second point was that the issue before the courts was whether the defendant, at the time of the crime, possessed a certain state of mind (under the M’Naghten rules that formalized the matter in1843, this concerned whether the defendant knew what he or she was doing at the time of the crime). While medical evidence about illness did indeed at times help the court decide reach a judgment, the establishment of the presence of illness, even psychotic illness, did not in itself require a finding of non-culpability. As medical witnesses began to understand and accept these points, doctors began to specialize in this kind of work and to acquire their own professional expertise. This was the beginning of the specialty of forensic psychiatry. A new profession, along with new administrative procedures, turned legal decisions about the insanity defence into a regular and ordered process.

A range of contingent factors affected the decisions actually taken. Over time, medical consensus about the existence of certain forms of psychosis usually led courts, where medical opinion agreed in diagnosing psychosis, to transfer the accused to hospital before the court hearing or to accept the insanity defence (or diminished responsibility). Because the courts tried to give understandable evidence for decisions, this could look as if they took the illness itself to be the deciding factor, though, formally speaking, it was not. Controversy lessened with time, especially, in Europe, with the ending of capital punishment, and with the introduction of the verdict of diminished responsibility, a verdict of guilt but guilt to a degree lessened by illness (or, we might say, by constraint). There have been legal parallels to this in laws which make allowance for provocation, self-defence, crimes of passion, or simply actions with unintended consequences (as in manslaughter charges).

Nevertheless, every so often an exceptional case, a case where there is a large and emotional public interest, leads to renewed comment about what is being decided. When I first thought about the history of this area, the cases of Peter Sutcliffe (the English ‘Yorkshire ripper’ 1981) and of John Hinckley (the US would-be murderer of President Regan, in 1982) did this. The Hinckley case, indeed, led to major changes in Federal and State law. More recently, there is the exceptional case of Anders Behring Breivik. In the media, and it would appear even in the court itself, the decision about reaching a verdict was put as a decision about the finding of mental illness.Footnote 25 The court had the hugely symbolic role, as the institutional agent of Norwegian society as a collectivity, of deciding what to do in response to unprecedented violence aimed at the very principles of liberal community. The possibility of finding insanity had an important part in debate about what verdict would be socially and legally acceptable, but the court could not escape its socially assigned agency as the decision-making body. The decision was not only an empirical matter. Perhaps the awkwardness the court exhibited in its role points to why people keep commenting on the insanity defence: in the last analysis, there is no ‘right’ answer. In decisions that are not routine, there is always an element of ambiguity, the ambiguity of a social world in which people are compatibilists, building social relations on a language which mixes causes and intentions and in which social relations always raise the possibility of questions about responsibility. That is social life in conditions of liberal individualism.

All this lies in the background of the development of forensic psychology. A specialty so labelled developed in the past 50 years or so, though there was, in some jurisdictions (especially in the US), substantial interest in giving psychological evidence a place in the courts early in the twentieth century. The sort of evidence that psychologists had to offer was often about individual capacities or differences in responsiveness to circumstances. For example, psychologists gave evidence about the degree to which witnesses might be considered reliable observers, and psychologists presented evidence about whether there were direct causal links between watching pornographic violence and behaviour. Argument about such evidence did not directly discuss volition and causal determinism but rather sought persuasive, empirically grounded statements about the capacity, or psychological agency, of individuals. The evidence was specifically psychological and enriched description of what a person was thought to be able to do. It gave the courts more to go on in reaching decisions and in ensuring the public found decisions plausible. Some people have thought that neuropsychology should greatly enhance psychology’s place in the assessment of agency. But that aspiration has been overtaken by a social and political shift away from concern with assessment of agency (based on knowledge about the past), to assessment of risk (based on predictive knowledge). The courts are now turning to neuroscience evidence in the hope that it will enhance their capacity to estimate risk and hence decide what to do with offenders.

Whatever the utilitarian and administrative arguments for a system of governance based on predictable outcomes and on calculations of risk, rather than governance based on retrospective ascription of agency, contemporary western societies by and large maintain psychological accounts of agency as part of a discourse on responsibility. Jurisprudence has not thrown out the mens rea requirement, and public opinion still wants punishment for offenders. Everyday psychology continues to be much interested in the psychological states thought to accompany responsibility, and courts are expected to, and do, reflect this. How far this discourse will diminish, remains to be seen.

4 Conclusion

There is a substantial body of psychological knowledge that has enriched and rendered precise everyday discursive practices that continuously negotiate the attribution of agency to people. In psychological knowledge with an exclusively behaviourist or neuroscientific form, however, many people would judge the category of agency, as opposed to the category of cause, out of place. Agency is now commonly taken to be a psychosocial category and an attribute of people, or of selves, and, by analogy, institutions, and (logically) not something attributable to brains or bodies. The study of agency is the study of the way people assign, feel, and act on power in all its forms, from desire to governance, or to constraint on power, in their own lives and in the lives of others. The study of agency is hence also the study of how people attribute freedom, obligation, and responsibility. History clearly has a lot to say about this—all the more so as political, legal, and religious thought and practice developed the notion of the person as agent, and, analogously, of institutions as agencies, over many centuries, long before there was a specifically psychological notion of agency.

What I have just stated is muddied, however, by the fact that in earlier centuries ‘agency’ denoted the powers or capacities of states of existence like souls, gods, and material substances. And this usage, in which one might substitute the word ‘cause’ (or ‘prime cause’) for ‘agent’, continues. Thus, it is not linguistically incorrect to describe neurons, pharmaceuticals, emotions, intelligence, the genes, or whatever as agents. Similarly, one might refer to the agency of the will or of positive thought. But clearly, something different is meant when we refer to a person’s agency, and there is a body of opinion that would like to restrict the notion of agency to people—to psychosocial agency. All the same, what I am analytically distinguishing as two conceptions of agency often enough are merged in practice. This is yet another of many examples of the way psychology, taken in all its variety, has a hand in both natural science and in everyday forms of understanding people, and in explanations in terms of material causes and in explanations in terms of intentions, language, and mental processes.

Psychological discussions of agency are embedded in a web of historically formed meanings and power relations. Consider the spiritualist séance. In Victorian and Edwardian times, there was much debate about how to ascribe agency: to spirits, to hysterical women, to charlatans, to conjurors, to women seeking empowerment in a man’s world, to nervous reflexes or ideomotor action, to unconscious forces, to extrasensory perception, and so on.Footnote 26 It certainly matters to examine what went on and what was said about what went on in terms of participants’ and observers’ understanding of agency. The appropriateness of understanding agency in psychological terms at all (for example, attributing it to unconscious forces) was itself an issue. The most dismissive male doctors were inclined to attribute causation to unstable and even degenerate female bodies. The most ardent devotees were unshakeable in their belief that agency rested with spirits. Worldly sceptics attributed agency to the desire of people to make money. When historians write about the events, they too exercise agency in choosing which story to tell. For historians of psychology, the prime story is the place such phenomena had in the emergence of twentieth-century theories of suggestive influence and of the unconscious. A feminist might think differently.

As the history of the spiritualist séance suggests, the history of notions of psychosocial agency is inseparable from the history of notions of the self. I have used this essay to focus on the notion of volition, rather than on the self, however, and then to use medicolegal discussion to open up a more specific account, relevant to agency, of relations between forms of knowledge and social practice. The history of agency, understood psychosocially, is a history of claims about what individual people could and could not do, and hence of claims about what people could and could not be held responsible for. As early forms of character assessment like physiognomy and phrenology suggest, as fascination with hypnotic and séance states confirms, and as studies of late nineteenth-century and twentieth-century popular psychology bear witness, psychological work was intimately bound up with hopes and fears about what people could or could not do and how individual capacities could or could not change (Hayward 2014; Rose 1989/1999; Thomson 2006). The history of many modern psychological practices, or technologies, is a history of the modern culture of agency—in the work of therapy and counselling, and in the literature and training practices of self-help, memory improvement, acquisition of kills, bolstering of assertiveness, positive thinking, and so on. We may understand academic social psychology as having sought to give these practices a basis in systematic knowledge of interpersonal relations.

Earlier moral philosophers and psychologists, when they assigned agency to people, referred to theoretical entities like states of mind, intentions, volition , and free will , and more recently traits and personality. In the course of the twentieth century, recourse to such entities became a source of division among psychologists. There have been psychologists who decry such references—most recently and emphatically the eliminativist neuropsychologists who held that it was the task of science to translate everyday language about mind (disparagingly called ‘folk psychology’) into the language of neural events. Needless to say, there were many commentators, psychologists among them, who thought this completely wrong.Footnote 27 Moreover, many critics thought a non-eliminativist stance underpinned the ethical principle of respecting people’s agency.

All this activity and debate is characteristic of psychological society. As a form of social order, such society has many proponents, both lay and professional, and many detractors. Evaluation often turns on the question of agency. Awareness that agency is a social status, a status ascribed according to rules regulating power, leads critics of psychological society to claim that psychological thought individualizes notions of power and thereby hides its true nature, which in the political worlds we have is a function of socially inequitable structures. By contrast, practicing psychologists—and we may include both those who claim knowledge or skill and those without formal qualifications who seek knowledge or skill—perform with the conviction that individuals can achieve meaningful agency. This exhibits a driving assumption of political individualism: meaningful power and meaningful change exist at the level of the individual. Nevertheless, there are critical psychologists, such as K. J. Gergen, who argue that only due recognition of the social constitution of individual agency will free psychological practice to contribute to rather than diminish individual human agency .Footnote 28 Social psychologists, I think, have acknowledged something along these lines, insofar as their research has turned towards the ways in which people understand, ascribe, and negotiate the terms of agency in social relations.

The discourse of specifically personal agency manifestly has high value in contemporary liberal democracies. Linguistic usage going back to earlier centuries, however, makes it possible to describe the body, the brain, the unconscious, the social environment, economic forces, or indeed spirits, God, and devils as agents. Thus, the question, where to assign agency, and to what, places psychology at the centre of social and political debate about the power of the individual.