Attributing Agency to Automated Systems: Reflections on Human–Robot Collaborations and Responsibility-Loci

Many ethicists writing about automated systems (e.g. self-driving cars and autonomous weapons systems) attribute agency to these systems. Not only that; they seemingly attribute an autonomous or independent form of agency to these machines. This leads some ethicists to worry about responsibility-gaps and retribution-gaps in cases where automated systems harm or kill human beings. In this paper, I consider what sorts of agency it makes sense to attribute to most current forms of automated systems, in particular automated cars and military robots. I argue that whereas it indeed makes sense to attribute different forms of fairly sophisticated agency to these machines, we ought not to regard them as acting on their own, independently of any human beings. Rather, the right way to understand the agency exercised by these machines is in terms of human–robot collaborations, where the humans involved initiate, supervise, and manage the agency of their robotic collaborators. This means, I argue, that there is much less room for justified worries about responsibility-gaps and retribution-gaps than many ethicists think.     

Responsibility,Autonomy, and the Zygote Argument

In this paper I argue that the distinction between moral responsibility and autonomy can illuminate various debates about the Zygote Argument (as well as the Four-Case argument). Having made this distinction, one can see how these (broadly speaking) manipulation arguments are unsuccessful. Building on previous work, I also argue that this distinction can provide a framework for understanding other important work in agency theory, including that of Harry Frankfurt and Gary Watson.     

The Trouble with Externalist Compatibilist Autonomy

In this paper, I try to show that externalist compatibilism in the debate on personal autonomy and manipulated freedom is as yet untenable. I will argue that Alfred R. Mele’s paradigmatic, history-sensitive externalism about psychological autonomy in general and autonomous deliberation in particular faces an insurmountable problem: it cannot satisfy the crucial condition of adequacy “H” for externalist theories that I formulate in the text. Specifically, I will argue that, contrary to first appearances, externalist compatibilism does not resolve the CNC manipulation problem. After briefly reflecting on the present status of responses to the manipulation problem in the debate between compatibilists and incompatibilists of various stripes, I will draw the over-all pessimistic conclusion that no party deals with this problem satisfactorily.     

Setting the bar for cognitive agency: Or,how minimally autonomous can an autonomous agent be?

Embodied cognition comes in many variants. Yet, in most if not all of these variants the notion of agency, or more precisely autonomous agency, occupies center stage. However, whereas in its original context of application autonomy applies strictly to the human sphere, cognitive theory must needs generalize this concept so as to render it applicable to a much wider range of entities and processes. Theorists of embodied cognition must therefore strive to articulate a valid sense of minimal autonomy applicable to animals and, arguably, to artificial agents as well. In this paper I discuss two major attempts to articulate a conception of minimal autonomy which I describe, respectively, as the adaptive-behavior approach (ABA) and the systemic-constructivist approach (SCA). The differences between these two leading views on minimal autonomy reveal fundamental disagreements not only with respect to the nature of autonomous agency but also with respect to embodiment and the relevance of biological life to the making of mental life.     

Do corporations have minds of their own?

Corporations have often been taken to be the paradigm of an organization whose agency is autonomous from that of the successive waves of people who occupy the pattern of roles that define its structure, which licenses saying that the corporation has attitudes, interests, goals, and beliefs which are not those of the role occupants. In this essay, I sketch a deflationary account of agency-discourse about corporations. I identify institutional roles with a special type of status function, a status role, in which the collectively accepted function is expressed in part through its occupier’s intentional expression of her agency in that role (where the occupier is part of the group whose collective acceptance underwrites her having the relevant function in social transactions). I identify institutions as systems of status roles and show how this is compatible with seeing the agency of institutions generally, even over time periods in which there is complete change in role occupiers, as a matter of the contributions only of individual agents. I explain how the reduction of the institution to its members is compatible with its potentially having had a completely different membership. I show in the case of the corporation in particular that, once we see its origins and function, the surface features of legal discourse about corporate agency are misleading and are compatible with a deflationary account of corporate agency. I show in connection with this that the corporation is to be identified with its shareholders, and that where a corporation separates ownership and control, its managers and employees are proxy agents of the shareholders doing business under the corporate form. Finally, I canvass the legitimate ways of construing ordinary talk about corporate intention, belief, and so on, in light of this, none of which support the attribution of genuine agency or intentionality to any group per se associated with the corporation.     

Relational Autonomy and Paternalistic Interventions

Relational conceptions of autonomy attempt to take into account the social aspects of autonomous agency. Those views that incorporate not merely causally, but constitutively necessary relational conditions, incorporate a condition that has the form:

(RelAgency) A necessary condition for autonomous agency is that the agent stands in social relations S.

I argue that any account that incorporates such a condition (irrespective of how the relations, S, are spelt out) cannot play one of autonomy’s key normative roles: identifying those agents who ought to be protected from (hard) paternalistic intervention. I argue, against objections from Oshana, that there are good reasons for maintaining the notion of autonomy in this role, and thus that such relational conceptions should not be accepted. This rejection goes beyond that from John Christman, which holds only for those relational conditions which are value-laden.     

Plausibility,Manipulation, and Fischer and Ravizza

The manipulation argument poses a significant challenge for any adequate compatibilist theory of agency. The argument maintains that there is no relevant difference between actions or pro‐attitudes that are induced by nefarious neurosurgeons, God, or (and this is the important point) natural causes. Therefore, if manipulation is thought to undermine moral responsibility, then so also ought causal determinism. In this paper, I will attempt to bolster the plausibility of John Martin Fischer and Mark Ravizza's semicompatibilist theory of moral responsibility by demonstrating how their account provides a distinctive line of response to three important types of manipulation.     

Dissociative Identity Disorder,Ambivalence, and Responsibility

If someone with dissociative identity disorder (DID) commits a wrongful act, is she responsible? If one adopts the Multiple Persons Thesis, it may seem that one alter cannot be responsible for the actions of another alter. Conversely, if one regards the subject as a single person, it may seem that she is responsible for any actions she performs. I will argue that this subject is a single person, but one who suffers from delusions of disownership and therefore does not fulfill ordinary requirements for responsible agency. This is because she suffers from extreme ambivalence: her deep‐seated needs and desires conflict, and she forms alter‐personalities as a way to cope with inner discord without abandoning any of these contradictory impulses. However, although the ability to exercise autonomous agency is eroded in such cases, the capacity for autonomous agency is preserved. The subject with DID is weakly responsible for her wrongful acts.     

A relational account of intellectual autonomy

According to relational views of autonomy, some social relations or forms of dependence are necessary for autonomous agency. Recent relational theorists have primarily focused on autonomy of action or practical autonomy, and the result has been a shift away from individualistic conceptions of autonomy in the practical realm. Despite these trends, individualistic conceptions are still the default when it comes to autonomy of belief or intellectual autonomy. In this paper, I argue for a relational account of intellectual autonomy. Specifically, I claim that intellectual autonomy requires a sense of one’s standing as an equal member of the epistemic community.     

Water Crisis Adaptation: Defending a Strong Right Against Displacement from the Home

This essay defends a strong right against displacement as part of a basic individual right to secure access to one’s home. The analysis is purposefully situated within the difficult context of climate change adaptation policies. Under increasing environmental pressures, especially regarding water security, there are weighty reasons motivating the forced displacement of persons—to safeguard water resources or prevent water-related disasters. Even in these pressing circumstances, I argue, individuals have weighty rights to secure access to their homes. I explain how the home provides a functional context for conditions of autonomous agency. Being coerced from the home disrupts and subverts the conditions necessary for autonomous processes. I conclude by suggesting that the right to the home could be a foundational element of territorial rights.     

Biases and Heuristics in Decision Making and Their Impact on Autonomy

Cognitive scientists have identified a wide range of biases and heuristics in human decision making over the past few decades. Only recently have bioethicists begun to think seriously about the implications of these findings for topics such as agency, autonomy, and consent. This article aims to provide an overview of biases and heuristics that have been identified and a framework in which to think comprehensively about the impact of them on the exercise of autonomous decision making. I analyze the impact that these biases and heuristics have on the following dimensions of autonomy: understanding, intentionality, absence of alienating or controlling influence, and match between formally autonomous preferences or decisions and actual choices or actions.     

Extended cognition,personal responsibility,and relational autonomy

The Hypothesis of Extended Cognition (HEC)—that many cognitive processes are carried out by a hybrid coalition of neural, bodily and environmental factors—entails that the intentional states that are reasons for action might best be ascribed to wider entities of which individual persons are only parts. I look at different kinds of extended cognition and agency, exploring their consequences for concerns about the moral agency and personal responsibility of such extended entities. Can extended entities be moral agents and bear responsibility for actions, in addition to or in place of the individuals typically held responsible? What does it mean to be “autonomous” when one’s cognition is influenced and supported by a milieu of environmental factors? To answer these questions, I explore strong parallels between HEC’s critique of individualism in cognition, and feminist critiques of individualist accounts of self, agency, and autonomy. This relational and social conception of autonomous agency, as scaffolded and supported (or undermined and impaired) by a milieu of social, relational, and normative factors, has important lessons for HEC. Drawing together these two visions of distributed and decentralized aspects of personhood highlights how cognition, action, and responsibility are inextricably linked. It also encourages a reconceptualization of all cognition and all concerns about responsibility for actions, not simply as sometimes “extended” around individuals, but as fundamentally communal, social, and normative, with individual cognition and individual moral responsibility being derivative special cases, not the paradigm examples. Individuals are merely one of many possible loci of cognition, action, and responsibility.     

The Misguided Marriage of Responsibility and Autonomy

Much of the literature devoted to the topics of agent autonomy and agent responsibility suggests strong conceptual overlaps between the two, although few explore these overlaps explicitly. Beliefs of this sort are commonplace, but they mistakenly conflate the global state of being autonomous with the local condition of acting autonomously or exhibiting autonomy in respect to some act or decision. Because the latter, local phenomenon of autonomy seems closely tied to the condition of being responsible for an act, we tend to think of the former, global phenomenon as a condition of responsibility as well. But one can act autonomously, or manifest autonomy with respect to some occurrent state, without satisfying the conditions for autonomous agency. Autonomous agency and responsible agency are logically distinct in part due to the varient conceptions of rationality each calls for. Both agent responsibility and holding a person responsible imply a fairly ``thick' form of rationality, where rationality embodies a normative component and is a matter of satisfying criteria that are objective in the sense that they are independent of what a person happens to want or to value. But autonomous agency calls for a quite different, ``thin' conception of instrumental rationality.     

Autonomous agency,we-agency,and social oppression

Theories of collective intentionality and theories of relational autonomy share a common interest in analyzing the social dynamics of agency. However, whereas theories of collective intentionality conceive of social groups primarily as intentional and voluntarily willed, theories of relational autonomy claim that autonomous agency is both scaffolded and constrained by social forces and structures, including the constraints imposed by nonvoluntary group membership. The question raised by this difference in view is whether social theorizing that overlooks the effects of nonvoluntary social group membership on individual and joint agency overlooks crucial aspects of the social dynamics of agency. To explore this question, this article first evaluates Michael Bratman's planning analysis of individual agency from the perspective of relational autonomy theory and compares it with a narrative self-constitution account of temporally extended agency. It then evaluates Bratman's analysis of shared agency and discusses Shaun Gallagher and Deborah Tollefsen's concept of we-narratives, which extends the notion of narrative construction to shared agency. Overall, the argument aims to show that if we are interested in understanding the social dynamics of agency, it is critical to attend to the way that agents exercise their intentional agency in relation to internalized and external social constraints.     

On the Importance of History for Responsible Agency

In this article I propose a resolution to the history issue for responsible agency, given a moderate revisionist approach to responsibility. Roughly, moderate revisionism is the view that a plausible and normatively adequate theory of responsibility will require principled departures from commonsense thinking. The history issue is whether morally responsible agency – that is, whether an agent is an apt target of our responsibility-characteristic practices and attitudes – is an essentially historical notion. Some have maintained that responsible agents must have particular sorts of histories, others have argued that no such history is required. Resolution of this contentious issue is connected to a wide range of concerns, including the significance and culpability of different forms of manipulation, the plausibility of important incompatibilist criticisms of compatibilism, and of course, a satisfactory account of moral responsibility. As it turns out, history matters sometimes, but less frequently than we might think.     

Goal striving within agentic and communal roles: separate but functionally similar pathways to enhanced well-being

ABSTRACT Do agency and communion strivings provide functionally similar but predictively independent pathways to enhanced well‐being? We tested this idea via a year‐long study of 493 diverse community adults. Our process model, based on self‐determination and motive disposition theories, fit the data well. First, the need for achievement predicted initial autonomous motivation for agentic (work and school) role‐goals and the need for intimacy predicted felt autonomy for communal (relationship and parenting) goals. For both agentic and communal goals, autonomous motivation predicted corresponding initial expectancies that predicted later goal attainment. Finally, each type of attainment predicted improved adjustment or role‐satisfaction over the year. Besides being similar across agency and communion, the model was also similar across race and gender, except that the beneficial effects of communal goal attainment were stronger for high need for intimacy women and Blacks. Implications for agency/communion theories, motivation theories, and theories of well‐being are discussed.     

Supporting intimates on faith

What is the role of faith in the familiar practice of supporting intimates in their personal projects? Is there anything distinctly valuable about such faith-based support? I argue that the virtue of being supportive, a characteristic of the good friend or lover, involves a distinctive kind of faith: faith in another persons’ chosen self-expressive pursuit(s). Support based on such faith enables the supported party to enjoy a more meaningful and autonomous exercise of agency in self-expressive arenas, and engenders a sense of relational unity or solidarity, deepening the normative and emotional bonds of the relationship.     

Approaching Complex Cases With a Crisis Intervention Model and Teamwork: A Commentary

This article is a response to the case of SA presented by O'Daniel and Wells. In this response I support the approach used with the patient, and discuss and expand on the applicability of a crisis intervention model to genetic counseling. In addition, I explore the issue of autonomy and informed decision making for patients who are in crisis. I also discuss the surface contradiction between the mandate for nondirectiveness and the need to provide direction during crisis intervention are discussed and reframe the issue as one of process versus content directiveness. I argue that during times of crisis, genetic counselors may need to direct process in order for patients to be able to make autonomous decisions.     

Multiple aspects in the sense of agency

Recent significant research in a number of disciplines centers on the concept of the sense of agency. Because many of these studies cut across disciplinary lines there is good reason to seek a clear consensus on what ‘sense of agency’ means. In this paper I indicate some complexities that this consensus might have to deal with. I also highlight an important phenomenological distinction that needs to be considered in any discussion of the sense of agency, regardless of how it gets defined. Finally, I suggest that the sense of agency has an ambiguous phenomenology and I offer some critical comments on current models that fail to notice this ambiguity.