What’s Wrong with Motive Manipulation? Eric M. Cave Accepted: 27 October 2006 / Published online: 21 December 2006 # Springer Science + Business Media B.V. 2006 Abstract Consider manipulation in which one agent, avoiding force, threat, or fraud mobilizes some non-concern motive of another so as to induce this other to behave or move differently than she would otherwise have behaved or moved, given her circumstances and her initial ranking of concerns. As an instance, imagine that I get us to miss the opening of a play that I have grudgingly agreed to attend by engaging your sublimated compulsive tendency to check the stove when we are halfway to the theatre. Such motive manipulation is, I take it, widely regarded as morally worrisome. If it really is morally worrisome, then we should be able to explain adequately why it is so. But existing condemnations of manipulation come up short in this regard. In this paper, I develop and defend a more plausible account of the moral status of this phenomenon. Key words autonomy . bypassconceptionof autonomy . historical conceptionof autonomy . manipulation . motive manipulation Manipulation comes in many guises. One agent might manipulate another by means of deception. She might, for instance, tell another falsely that she needs gas money to get her small children home from the county fair. Or a manipulator might employ a threat that does not rise to the level of coercion. She might, for instance, seek to seduce another by telling him that she will cease to be his friend unless he acquiesces to sex immediately. Or an agent might manipulate another by artificially constraining this other’s options. She might, for instance, publicize another’s secret vow to stop smoking to make it harder for him to bum cigarettes. I could go on, for variants of manipulation are legion. Perhaps one agent can manipulate another without intending to do so. But morally speaking, the most interesting cases of manipulation are intentional, not accidental. In what follows, I shall focus on a subset of such cases, understanding “manipulation” accordingly. Ethic Theory Moral Prac (2007) 10:129–144 DOI 10.1007/s10677-006-9052-4 E. M. Cave (*) Department of English and Philosophy, Arkansas State University, P.O. Box 1890, Jonesboro, AR 72467, USA e-mail: ecave@astate.edu

Perhaps it is possible to delineate a set of necessary and sufficient conditions that mark off manipulation from other human behaviors. I am not optimistic about the prospects of so doing, however, because I suspect that “manipulation” refers to a number of different phenomena, not all of which overlap in their interesting features. Accordingly, I shall begin not from a general analysis of manipulation, but by narrowing to the particular phenomenon that I intend to explore. As a preliminary matter, consider motives. I shall understand motives in a very broad way, as all of the states of an agent capable of causing behavior or movement on the part of that agent. An agent’s motives, so understood, are a mixed bag, including both psychological and physiological states. Desires are motives, on this understanding, as are preferences and intentions. Psychological drives are motives as well, such as the drive to create art. Physiological drives, such as hunger and thirst, are motives, too. Even physical reflexes, such as flinching away from sudden and unexpected movement, count as motives in this broad sense. Why understand motives so broadly? My inquiry here is about manipulation, and any of the states capable of causing an agent to move can be used to manipulate that agent. Obviously, agents can be manipulated by means of their desires, preferences and intentions. Any competent interrogator will realize that agents can be manipulated by means of their psychological and physiological drives as well. Even an agent’s physical reflexes can be used to manipulate her. As an instance, consider the teenager who asks his date to a really scary movie in hopes that she will startle repeatedly into his arms. A broad understanding of motives may pick out a miscellany, but it is a miscellany unified by its potential usefulness to manipulators. Consider a distinction among an agent’s motives, so construed. Some motives are concerns. To be a concern, a motive must be a pro- or con-attitude taken by an agent towards a possible outcome of an available behavior or movement, and it must be accessible to that agent’s consciousness. Thus, conscious desires are concerns, as are conscious preferences and intentions. What remains of an agent’s motives after separating out her concerns are her non-concern motives. A non-concern motive is any state of an agent capable of causing her to behave or move which is either not a pro- or con-attitude taken towards a possible outcome of an available behavior or movement, or not accessible to her consciousness. Thus, an agent’s non-concern motives may include such things as desires, preferences, and intentions entirely masked by self-deception, psychological and physiological drives of which an agent is aware but towards which she is indifferent, and physical reflexes, just to name a few possibilities. On a classical rational choice approach, non-concern motives cause concerns, and concerns cause actions. But one can manipulate another not just by causing her to act, in the strong sense of intentionally undertaking to do something underwritten by the apparent balance or reasons, but also by causing her to behave or move in certain ways. Thus, an inquiry about manipulation requires us to think broadly about the relationship between non- concern motives, concerns, and behavior, at least relative to classical rational choice theory. As I shall conceive of them, non-concern motives can cause concerns. My subconscious desire for acceptance, for instance, can cause me to take up a conscious pro-attitude towards membership in a particular church. Or I might develop a concern for you when our being forced together at work assuages a subconscious desire of mine for human company. But concerns can cause non-concern motives as well. My concern to be the perfect parent/child can cause me to develop rebellious urges of which I am not aware. 130 E.M. Cave

Once they are in place, both concerns and non-concern motives can directly cause behavior or movement in an agent. My diving under the desk might be caused by a desire to hide from the person just entering the room, or it might be caused by a passing train’s activating a reflex developed during my time in earthquake-prone Southern California. Another illustration: my failure to greet you might be caused by a conscious desire of mine to snub you, or by a subconscious desire of mine to avoid acknowledging your existence because you resemble someone I dislike. In some recent discussions of manipulation, considerable attention is paid to whether motives appealed to by manipulators are resistible or irresistible, and to whether a target could shed such motives or not did she attempt to do so.1 Perhaps the most striking cases of manipulation involve the mobilization of irresistible motives which cannot be shed. But in ordinary usage, I count as manipulation cases in which one agent mobilizes (in a way yet to be described) motives of another which are merely difficult to resist and to shed. Setting up an attempt to characterize a kind of manipulation in a way that respects ordinary usage, I shall conceive of both concerns and non-concern motives as a mixed bag. Some are resistible, some irresistible. Some are able to be shed by the agent whose concerns they are, some are not. Concerns, as I have described them, are accessible to consciousness. This is not to say that they are transparent to the agents whose concerns they are. Rather, a concern is accessible to consciousness if an agent could discover it, with more or less accuracy, were she to reflect ex post upon past behaviors or movements, and ex ante upon possible future behaviors or movements, privileging the former over the latter when they disagree. Notice that accessibility to consciousness does not mark the divide between concerns and non- concern motives. While all concerns are accessible to consciousness, some non-concern motives are as well. Again, to be a concern, a motive must be both accessible to consciousness and a pro- or con-attitude. Any motive failing to satisfy either of these conjuncts is a non-concern motive. Obviously the line between concerns and non-concern motives is not hard and fast. Enhanced powers of discernment may transform what was once one of an agent’s non- concern motives into one of her concerns, and what may be a concern of a discerning agent may be a non-concern motive of a less discerning agent with similar dispositions to behave and to move. But this agent-relative aspect of the distinction between concerns and non- concern motives is not a shortcoming. Ultimately, it helps to insure that whether a particular effort counts as manipulative or not depends in part upon features of the psychology of the target of this effort. And this is just what we should aim for in characterizing any variant of manipulation, at least if we hope to respect ordinary usage of the concept. I shall assume that however agents access their concerns, they can order them according to the strength of the pro- and con-attitudes constituting them. And I shall assume that orderings of concerns can vary in completeness and coherence from one agent to another. An agent need not have complete concerns to be manipulated as I shall describe. She need only have pro- and con-attitudes that target the possible outcomes of the behaviors and movements available in her current situation. And an agent need not have fully coherent 1 On the issue of resistible and irresistible motives, see Fischer (2004). On the issue of motives which can and cannot be shed, see Mele (2001), pp. 144–73. What’s wrong with motive manipulation? 131