Washington, N., and Kelly, D. (2014). ‘Should an Individual Composed of Selfish Goals be Held Responsible for Her Actions?’ commentary on Julie Y. Huang and John A. Bargh “The Selfish Goal: Autonomously Operating Motivational Structures as the Proximate Cause of Human Judgment and Behavior,” Behavioral and Brain Sciences, 37(2): 158-159.

See full article with commentary and author response here 

By Natalia Washington and Daniel Kelly

Word Counts:
Abstract: 58 words
Main Text: 965 words
References: 294 words
Total Text: 1337 words

Should an Individual Composed of Selfish Goals be Held Responsible for her Actions?

Abstract: We discuss the implications of the Selfish Goal model for moral responsibility, arguing it suggests a form of skepticism we call the locus problem. In denying that individuals contain any genuine psychological core of information processing, the model denies the kind of locus of control intuitively presupposed by ascriptions of responsibility. We briefly consider ways the problem might be overcome.

Recent work has sparked various forms of the worry that the facts of human cognitive makeup lead to skepticism about free will, moral responsibility, and whether reflective deliberation and conscious decisions actually influence behavior. (Nahmias (2010), Roskies (2006), Doris (forthcoming), Levy (2012), Saul (forthcoming)). Rather than contest the substance or details of the target article, we assume for discussion that it is on the right track, and draw out one implication for common practices of holding our selves and each other responsible for our behaviors. We argue that the picture of an individual person as fragmented into a cluster of autonomously operating selfish goals competing for psychological resources suggests an interesting kind of problem for responsibility ascriptions.

Huang and Bargh marshal an impressive amount of empirical evidence in support of their Selfish Goal model, which they articulate in terms of four principles (autonomy, reconfiguration, similarity, and inconsistency). For our purposes, two features are most important, and both focus on control and behavior. First, while the model depicts an individual person as psychologically fragmented (rather than integrated or unified), it does not suggest that the individual’s actions will be utterly random or under no control whatsoever. Nor does it suggest that those actions will be completely detached from internal psychological processes. Rather, a specific episode of behavior will ultimately be under the control of a specific goal, namely whatever goal has become active and won the competition for access to the lower level machinery closer to the behavioral periphery. Importantly, goals, on Huang and Bargh’s view, are where the buck stops; they are the highest level, most sophisticated psychological states that enter into the production of overt behavior. There is no higher court to appeal to, no further, more global mental entity that selects among the goals, unifies the individual psychologically, or serves as a more constant source of control over different goal-driven episodes of behavior (hence the propensity for inconsistency that they stress). Second, while they do not deny the existence of a self, they allow it only a severely limited function, relegating it to the politician or public relations role of constructing rationalizations for actions that the self has no hand in selecting, producing, or controlling. With the exception of linguistic behavior, the self is behaviorally epiphenomenal. (target article 8, 56, 57)

This picture appears to be at odds with much everyday thought, including thinking about moral responsibility. According to common practices, an individual is responsible only for those behaviors over which she has proper control (as opposed to behaviors that are coerced, accidental, or brutely reflex-like). When she successfully wields such control, the resulting behavior is an appropriate target for responsibility ascriptions. While there is not yet any clear consensus on the exact form of control required, a recognizable theme is that individuals contain a stable, continuous psychological core that is the source of this control when it is properly exercised. That is, control is a two-part relation, with one relata being the behavior, and the other we will call the psychological locus. With the term “locus” we are trying to remain neutral on whether one thinks of this psychological feature in terms of an individual’s self, character, identity, Cartesian mental substance, immortal soul, or whatever. However it is cashed out, typical responsibility ascriptions presuppose that individuals have this kind of central psychological feature. In other words, in ascribing responsibility for an action, there is something we are ascribing responsibility to—some central element of the individual that is the ultimate source of the behavior, and to which responsibility attaches.

This then is the worry: the Selfish Goal picture suggests that individuals contain no such psychological locus, because our behaviors are ultimately determined by a loose collection of autonomously operating goals, each with its own agenda. From the perspective of moral responsibility, neither praise nor blame for behavior will be justified because psychologically there’s no ‘there’ there.

We do not think the locus problem is insurmountable. Even if the Selfish Goal model is correct in denying the existence of a traditional psychological locus, there are reasons to think individuals do have the tools to act as responsible agents. An increasingly prominent movement in philosophy of mind emphasizes the embodied, distributed, social, and externalized character of much cognition. Thinkers like Dennett (2003) and Clark (2008) stress that even sophisticated human behaviors can be guided by decentralized control systems that lack a stable or continuous core (see also Shapiro 2007). Two recent elaborations on these themes deal more directly with moral responsibility and agency.

Doris’ (forthcoming) dialogic conception of agency recasts the post hoc and socially mediated rationalizations that are the domain of Huang and Bargh’s ‘conscious self’ as in part constitutive of human agency. For an individual to participate in this kind of moral reasoning allows that individual’s behaviors to become self-directed, or guided by those goals relevant to her most important values. Thus an individual can justifiably be held responsible for a behavior even if that behavior is not under the control of, say, her internal locus of reflective deliberation.

Ismael (2007, 2010) argues that selves should be thought of in terms of self-governing systems. For her, the conscious self is not merely epiphenomenal, but part of a feedback loop that evolved to help organize and guide sophisticated behaviors. Here, the self is not a centralized controller, but a distributed but higher level subsystem of the mind “perfectly compatible with a fully decentralized understanding of lower-level processing”

Both projects point to conceptions of agency and behavior control that do not rely on the existence of a traditional, stable, centralized psychological core. If there is such thing as a ‘real self’ on these views, it is a dynamic, distributed sort of thing. While it doesn’t guarantee complete consistency in our behavior, but it may get us what responsibility ascriptions seem to require. By our lights, both look well suited to accommodate both our every day practices of holding individuals responsible, and the Selfish Goal’s denial of a psychological locus.


Clark, A. (2008). “Soft Selves and Ecological Control,” in D. Spurrett, D. Ross, H. Kincaid and L. Stephens (eds) Distributed Cognition and the Will. Cambridge, MA: The MIT Press.
Dennett, D. (2003). Freedom Evolves. New York: Penguin Books.
Doris, J. (Forthcoming). Talking To Ourselves. New York: Oxford University Press.
Ismael, J. (2007). The Situated Self. New York: Oxford University Press.
Ismael, J. (2011). “Self-Organization and Self-Governance,” Philosophy of the Social Sciences, 41(3): 327-351.
Levy, N. (2012). “Consciousness, Implicit Attitudes, and Moral Responsibility,” Nos, doi: 10.1111/j.1468-0068.2011.00853.x.
Nahmias, E. (2010). “Scientific Challenges to Free Will,” in T. O’Connor and C. Sandis (eds) A Companion to the Philosophy of Action. New York: Wiley-Blackwell.
Roskies, A. (2006). “Neuroscientific Challenges to Free Will and Responsibility,” Trends in Cognitive Science, 10(9): 419-423.
Saul, J. (Forthcoming.) “Implicit Bias, Stereotype Threat and Women in Philosophy.” Women in Philosophy: What Needs to Change?, Edited by Fiona Jenkins and Katrina Hutchison, New York: Oxford University Press.
Shapiro, L. (2007). “The Embodied Cognition Research Programme,” Philosophy Compass 2(2): 338-346.

Comments, suggestions, clever derision and unqualified flattery all welcome at