Social Science - Philosophy Assignment 272

Manipulation: Theory and Practice

Christian Coons (ed.), Michael Weber (ed.)

https://doi.org/10.1093/acprof:oso/9780199338207.001.0001

Published: 2014 Online ISBN: 9780190228446 Print ISBN: 9780199338207

Search in this book

CHAPTER

https://doi.org/10.1093/acprof:oso/9780199338207.003.0006 Pages 121–134

Published: August 2014

Abstract

Keywords: manipulation, influence, choice architecture, autonomy, libertarian paternalism, nudging, ethics

Subject: Social and Political Philosophy, Moral Philosophy

Collection: Oxford Scholarship Online

5 A Framework for Assessing the Moral Status of “Manipulation”  J.S. Blumenthal-Barby

This chapter addresses the question of the wrong-making features of manipulation and the

circumstances under which it is ethically justi�ed. It argues that it is not the case that any of the

practices traditionally labeled as “manipulation” are ipso facto morally wrong, nor is it even the case

that any of these practices have a single wrong-making feature (e.g., infringement on autonomy that is

always present but may be outweighed by other morally relevant factors). The moral status depends on

the extent to which an instance of in�uence (1) threatens or promotes autonomy, (2) has good aims

and virtuous overtones or bad ones, and (3) ful�lls or fails to ful�ll duties, obligations, and

expectations that arise out of the relationship between the in�uencer and in�uenced. The chapter

analyzes these domains in depth and discusses the challenges arising from the application of a

pluralistic framework such as this.

THIS ESSAY DEALS with the ethics of using knowledge about a person’s particular psychological make-up, or

about the psychology of judgment and decision making in general, to shape that person’s decisions and

behaviors. Various moral concerns emerge about this practice, but one of the more elusive and

underdeveloped concerns is the charge of manipulation. It is this concern that is the focus of this essay.

The elusiveness and underdevelopment of the manipulation charge goes something like this: Person A

describes a decision- or behavior-shaping practice (e.g., o�ering rewards, inducing a�ective states such as

guilt or fear, presenting information in a certain order or tone) and Person B claims “That’s manipulation,

and therefore morally wrong!” This claim is usually followed by Person A’s attempting to explain why

D ow

nloaded from https://academ

ic.oup.com /book/4870/chapter/147239951 by Florida State U

niversity user on 04 January 2024

instance X is not an instance of manipulation, and therefore morally permissible. This is the wrong

response, I would argue. The more illuminating response would be to consider what if anything is morally

problematic about manipulation. So, the discourse is underdeveloped in that it relies on intuitions that

manipulation is morally wrong instead of providing arguments as to why and how it is, and it is elusive in

that it is never quite clear on what exactly is meant by manipulation, descriptively speaking. For example, is

any sort of intentional in�uence without the subject’s knowledge manipulation? What about a case in which

the subject is aware of the in�uence attempt but has trouble resisting it because it plays on desires or other

parts of the subject’s psychological make-up? Is that a case of manipulation as well?

The literature has been less than clear on this conceptual point. The following practices have all been

described by philosophers as manipulation: incentivizing, o�ering, increasing options, decreasing

options, tricking, using (resistible) threats of punishment, managing information, presenting information

in a way that leads to predictable inferences, deceiving, lying, making a false promise, withholding

information or options, slanting information, providing irrelevant inputs or crowding out relevant inputs,

exaggerating information in a misleading way, using misleading packaging or misleading images, creating

impressions by imagery, using loaded language, trading on fear, subliminal suggestion, insinuations,

�attery, guilt, appealing to emotional weaknesses or needs, initiating psychological processes that are

di�cult to reverse or that lead to predictable behaviors or decisions (e.g., the tendency to continue with an

active decision even after becoming aware that it is more costly than originally thought, or the tendency to

view an option as more desirable when shown its contrast), browbeating or otherwise wearing the person

down, reverse psychology, and seduction. Even if we were to achieve some clarity on the bounds of

manipulation, doing so would fail to answer the central moral question, which concerns the moral status of

any of these practices.

p. 122

1

The moral status of any of these practices, I would argue, depends. That is, it is not the case that any of these

practices traditionally labeled as “manipulation” are ipso facto morally wrong; nor is it even the case that

any of these practices always has a single wrong-making feature (e.g., infringement on autonomy) that is

always present but may be outweighed by other morally relevant factors such that the practice may be all

things considered ethically permissible or morally right. What then does the moral status of these methods

of in�uence that lie somewhere between reason and coercion depend on? I argue that the moral status

depends on the extent to which the instance of in�uence (1) threatens or promotes autonomy; (2) has good

aims and virtuous overtones or bad ones; and (3) ful�lls or fails to ful�ll duties, obligations, and

expectations that arise out of the relationship between the in�uencer and in�uenced. I will explain in more

detail the moral relevance of these factors, showing why each is necessary and demonstrating how they

work in speci�c cases.

p. 123

Before doing so, a minor terminological note. Henceforth I will refer to the aforementioned practices not as

“manipulation” but as non-argumentative in�uence. Non-argumentative in�uence is in�uence that

operates either by bypassing a person’s awareness or by relying on facts about the subject’s psychology,

such as knowledge about his emotions, how he perceives things, how he makes judgments and decisions,

and what he desires. It is in contrast to in�uence that operates by engaging with a person’s rational

capacities and o�ering him reasons and arguments (i.e., rational persuasion), and to in�uence that operates

by force or severe threats of harm (i.e., coercion). In using the term “non-argumentative in�uence” instead

of “manipulation,” I hope to avoid any automatic negative associations and assumptions about moral

status and thereby address the central question of what if anything is ethically problematic about these

forms of in�uence and when.

2

D ow

nloaded from https://academ

ic.oup.com /book/4870/chapter/147239951 by Florida State U

niversity user on 04 January 2024

Autonomy

The majority of writing on the morality of non-argumentative in�uence is infused with condemnation of it.

The foundation of much of this is the claim that non-argumentative in�uence somehow interferes with an

agent’s autonomy. Melissa Seymour Fahmy condemns non-argumentative in�uence by arguing, “to

interfere in the lives of others without their consent is to usurp their authority to direct their lives as they

see �t.” Eric Cave makes a similar argument in previous work and in this volume, claiming that when A

induces B to behave di�erently than she otherwise would have via non-argumentative in�uence, A

undermines B’s capacity to manage her concerns since B cannot manage her concerns e�ectively while A is

doing so. Cave argues that non-argumentative in�uence violates “modest autonomy,” where modest

autonomy means that “To act autonomously, an agent need only act from concerns not thrust on her by

others in ways that overwhelm her capacity for control over her own mental life.” James Stacey Taylor

builds absence from non-argumentative in�uence into his de�nition of autonomy, arguing that a person is

autonomous with respect to a choice that she makes if and only if (i) the information on which she based the

choice has not been a�ected by another agent with the end of leading her to make a particular choice, or a

choice from a particular class of choices; or (ii) if it has, then she is aware of this. He writes, “It is clear that

the successful manipulation of a person into performing an action that she would not have otherwise

performed would serve to compromise the manipulee’s autonomy with respect to her manipulated

actions.” He recounts the story of Shakespeare’s Othello where Iago insinuates to Othello that his wife

Desdemona is having an a�air by arranging for Othello to overhear certain conversations and planting

suggestive objects such as a handkerchief until Othello becomes convinced and jealous enough and kills

Desdemona. Taylor writes, “Since Othello was manipulated by Iago into smothering Desdemona, he lacked

autonomy with respect to his decision to do this, for it was not he, but Iago, who was the font of this

decision. It was not Othello, but Iago, who originally decided to cause Desdemona’s death; Othello was

merely the instrument through which he brought it about.”

3

p. 124 4

5

6

7

In response to the claim that non-argumentative in�uence poses a threat to autonomy, Sarah Buss has

produced a compelling counter-argument. Buss’s argument consists of two main claims. First, she argues

that, at least on a compatibilist account, we consider people to be self-governing or autonomous despite the

fact that there is a past and an external world that in�uences their current mental states and actions, and

that non- argumentative in�uences are no di�erent; if the past and the external world do not pose threats

to self-governance or autonomy then neither does non-argumentative in�uence from others. Second, she

argues that many well-informed, self-governing agents would endorse a policy that involved being

in�uenced by non-argumentative forms of in�uence (e.g., seduction), and though she does not spell this

out, this sort of endorsement would result in the in�uence’s being compatible with autonomy under

accounts such as John Christman’s, where a person’s development of some desire is autonomous only if the

person “would not have resisted that development had [he] attended to the process [by which it was

developed].” Incidentally, Harry Frankfurt has made a point similar to Buss’s �rst claim when he writes

“We are inevitably fashioned and sustained, after all, by circumstances over which we have no control....It is

irrelevant whether those causes are operating by virtue of the natural forces that shape our environment or

whether they operate through the deliberatively manipulative design of other human agents.” For

Frankfurt, of course, what matters for autonomy is whether a person is wholeheartedly behind the desires

that move him to act, regardless of the origin of those desires.

8

p. 125

9

10

11

I think that Buss succeeds in countering the prevailing view that non- argumentative in�uence, or what has

traditionally been called manipulation, is always incompatible with autonomy. When then is it incompatible

with autonomy? Let me make a couple of modi�cations to the Buss–Frankfurt claims and then o�er some

criteria that allow us to assess the extent to which an instance of non-argumentative in�uence is

compatible with autonomy. The main modi�cation that I would make is to reject the equivalence of a non-

D ow

nloaded from https://academ

ic.oup.com /book/4870/chapter/147239951 by Florida State U

niversity user on 04 January 2024

argumentative in�uence that occurs undirected from the environment (e.g., a song that comes on the radio

and in�uences someone to behave romantically, or the natural absence of an important piece of

information that in�uences the direction of someone’s decision) and one that occurs directed at someone

from someone else (e.g., A knows about B’s weakness for jazz and so puts on a jazz CD to in�uence B to

behave romantically, or A withholds an important piece of information from B to get B to make a particular

decision). Why is this distinction important for autonomy? To see the importance we need only consider

why it is that we think that something like slavery is so morally egregious: it is not just because the slave is

not able to govern himself, it is because he is governed by someone else. The master has imposed his will on

the slave in a way that the slave would not endorse. And the second modi�cation I would make to the Buss

account is to emphasize the signi�cant di�erence between contexts of seduction and romance, which Buss

focuses her analysis and claim of endorsement on, and other contexts. It may be that most of us would

endorse, or at least not repudiate, being moved to love someone via non-argumentative in�uences such as

romantic gestures, but this sort of endorsement is not likely to be the case in many other instances of non-

argumentative in�uence, such as if I am moved to consent to a surgery for my elderly grandmother via the

non-argumentative in�uence of guilt used by my physician to induce me to consent.

p. 126

12

Thus, the criteria that I would o�er to allow us to assess the extent to which an instance of non-

argumentative in�uence is or is not compatible with autonomy are as follows:

1. The extent to which the non-argumentative in�uence attempt blocks or burdens options; and

2. The extent to which the person in�uenced is aware of and endorses, or were he were aware would

endorse, the non-argumentative in�uence attempt as a process by which his desires, decisions, or

actions were formed.

These are of course not new or novel criteria, as they stem largely from neo-Frankfurtian theories of

autonomy such as Christman’s, but it is worth explicitly outlining them as the key normative questions

that should be asked when considering the compatibility of non-argumentative in�uence and autonomy,

instead of assuming or tautologically holding that non-argumentative in�uence interferes with autonomy.

p. 127

Before continuing on to the other morally relevant factors and developing the rest of the framework,

however, it is worth pausing to address an important question and to make an important observation. The

important question is whether an agent whose autonomy is impaired by one instance of non-argumentative

in�uence can have his autonomy further impaired by another one. Consider Johnny who decides to start

smoking after much non-argumentative in�uence from tobacco companies, a process of desire formation

that he does not endorse upon re�ection. Thus, Johnny is not autonomous with respect to his desire to

smoke. Johnny’s physician decides to use counter non-argumentative in�uence—let us say extremely

frightening videos—to try to get Johnny to stop smoking. How is it even possible that the physician’s

actions could impair Johnny’s autonomy with respect to his smoking, given that he is already non-

autonomous with respect to that decision? In other words, if Johnny’s autonomy is at a zero, so to speak,

how could it go any lower? Investigating this in great depth would take us too far a�eld, but my intuition is

that one way in which it could is that this further manipulation (which, let us say, Johnny would not

endorse, either) makes it harder for Johnny to recover and govern himself, and in that sense does pose a

further threat to autonomy.

There is, as I mentioned, an important observation with respect to the impact of non-argumentative

in�uence on autonomy as well. Earlier I de�ned “non-argumentative in�uence” as in�uence that operates

either by bypassing a person’s awareness or by relying on facts about the subject’s psychology such as

knowledge about his emotions, how he perceives things, how he makes judgments and decisions, and what

he desires. It strikes me that the former type is the most threatening to autonomy. Before I elaborate, let me

make a brief conceptual point, and that is that the “or” is not meant to be an exclusive or. Often instances of

D ow

nloaded from https://academ

ic.oup.com /book/4870/chapter/147239951 by Florida State U

niversity user on 04 January 2024

non-argumentative in�uence that bypass a person’s awareness do so by relying on knowledge of facts

about their psychology (e.g., that he does not notice information that is placed in the middle instead of �rst

or last, or that he simply follows what is presented as the norm or the default without even realizing it). But

many instances of non-argumentative in�uence that rely on knowledge of facts about a person’s

psychology (e.g., what he desires, how he perceives things) do so in a way that the subject is very well aware

of (e.g., inducement of desire or guilt).

It strikes me that autonomy is more threatened by the types of non- argumentative in�uence that bypass an

agent’s awareness. After all, if autonomy involves governing oneself, this requires some degree of

awareness and control over one’s actions and reasons for acting. The importance of this is acknowledged in

most accounts of autonomy, including Christman’s, which is illustrated by his inclusion of the hypothetical,

“were she aware of the process of desire/belief/decision/action formation...would endorse.” One way to put

this is that most accounts of autonomy require some sort of “responsiveness to reasons,” whether those

reasons are found in the demands of some external normative framework or inside a person’s own desires

and preferences. Thus, a non-argumentative in�uence that bypasses a person’s awareness is one that gives

him little chance to endorse it or not endorse it as the process by which his decisions and behaviors are

formed. Of course, it is possible that the agent would meet the criteria to satisfy the hypothetical—that is,

he would endorse it—but epistemologically we have very little insight as to whether the person would

endorse this process that he is not aware of. Thus, the epistemic and justi�catory burden is placed on the

in�uencer who is bypassing the agent’s awareness to make the case that the agent would in fact endorse it

as a process by which his desires, decisions, and behaviors were formed and as such it is autonomy

preserving.

p. 128

Now, onto what I take to be the second morally relevant consideration for the moral status of non-

argumentative in�uence: whether it has good aims and virtuous overtones or bad ones.

Aims and Overtones (Virtuous or Vicious)

One objection that has been raised against non-argumentative in�uence is that the practice is often aimed

at bad ends or involves non-virtuous elements such as dishonesty or violation of trust, disrespect,

arrogance, predatoriness, and laziness. Laziness, because it takes more e�ort to sit down and try to

convince someone of a point or a course of action by laying out arguments and responding to objections and

questions than it does to appeal to his emotions or set up the environment a certain way. Predatoriness,

because, as Colin McGinn notes, the in�uencer studies human weakness and plays on anxieties and

insecurities; and as Marcia Baron notes, exploits emotional needs or a sense of indebtedness.

Arrogance, because of the hubris involved in thinking that I know better than you what is good for you and

that you are not capable of understanding what is best and why. This harkens back to Aristotle’s argument

that tools of rhetoric are to be used when dealing with “people whom one cannot instruct,” such that non-

argumentative in�uence becomes necessary when the quality of the audience degrades. And, disrespect,

because, as Fahmy claims, to use non-argumentative in�uence is to treat the subject as a child, and as

Robert Noggle claims, is to degrade her and treat her as less than a person and more like a machine whose

levers can be pushed and pulled. Marcia Baron, Allen Wood, and Kate Manne all make similar claims in this

volume about viewing others as mere play things, not taking them seriously, and being “too ready” to

manage them. And �nally, dishonesty or violation of trust, since non-argumentative in�uence involves

dealing with others in a less than straightforward way, which may damage interpersonal trust.

13p. 129 14

15

16

17

18

19

20

But as Baron points out in earlier work, non-argumentative in�uence can also involve virtue, and of course,

good aims. Baron writes, “the person who has the virtue corresponding to manipulativeness—a virtue for

which we do not, I believe, have a name—knows when it is appropriate to try to bring about a change in

D ow

nloaded from https://academ

ic.oup.com /book/4870/chapter/147239951 by Florida State U

niversity user on 04 January 2024

another’s conduct and does this for the right reasons, for the right ends, and only in instances where it is

warranted (and worth the risks), and only using acceptable means.” She cites the examples of persuading

a reluctant friend or spouse to seek medical help or prodding an untenured friend with a slim publication

record to produce. To not use non-argumentative in�uence at certain times or to adhere to a form of

fanatical straightforwardness is actually a vice, Baron argues, a vice of isolationism. I think that she is right.

Baron gives the example of not stopping a friend from driving home drunk. To not use a form of non-

argumentative in�uence, such as hiding the keys or distracting the friend by tempting him with a cup of his

favorite co�ee, is to exhibit the vice of isolationism. To quell potential dismissals of this example owing to

the subject’s intoxication, let me o�er another example. Imagine a person whose friend and roommate is

overweight and developing diabetes. The person creates an environment that non-argumentatively

in�uences the roommate to eat healthier by stocking the cupboards with healthy snacks and priming the

roommate to take walks by leaving running magazines and shoes out around the apartment. In these

cases, the use of non-argumentative in�uence, I would argue, involves the virtues of respect for the worth

of others, kindness, and courage. It is hard to say there is any dishonesty involved—after all, it is not the

case that the friend is putting distorting mirrors up around the apartment to get the roommate to see

himself as more overweight than he actually is. It is also hard to say that there is laziness involved. In fact, it

would be easier to present the friend with a simple argument full of statistics about why he should lose

weight. And it is also hard to say that there is any disrespectful, predatory, or arrogant behavior.

21

p. 130

22

Thus, whether practices of non-argumentative in�uence are ethically problematic via ends or virtue-based

concerns depends. Although usually associated with vicious traits such as disrespectful, predatory, and

arrogant, the use of “manipulation” or non-argumentative in�uence can aim at good ends and be

motivated from and exhibit virtuous traits such as kindness, courage, and respect for the worth of others.

Our intuitions about the extent to which an instance of non-argumentative in�uence exhibits virtue or vice

depends in part on the context in which the non-argumentative in�uence takes place or, to be more

speci�c, on the relationship between the in�uencer and the in�uenced, and the duties and boundaries that

arise out of that relationship. So, on to the next factor relevant to assessing the moral status of

interpersonal in�uence: the relationship between the in�uencer and the in�uenced, and the duties,

obligations, and expectations that arise out of that relationship.

Relationship Between Influencer and Influenced

Consider the di�erence between two cases: (1) the government begins priming people to eat healthier by an

advertising campaign that vividly illustrates the negative health and aesthetic consequences of obesity,

playing on both fear and ego; and (2) a wife begins priming her husband to eat healthier by reminding him

of his father who died young of a heart attack, playing on a fear she knows he has, and also occasionally

pointing out how attractive he looks after he has just exercised, playing on his ego. The latter case is one

that is familiar and usually considered unobjectionable. As a member of the audience where I presented this

paper remarked about the latter case, “What’s the big deal with that example of manipulation? My wife

knows a thousand psychological tricks to use on me, and it would be odd to say that my autonomy is

impaired or that something is morally problematic about what happens in such routine interactions.” His

remark serves as a reminder about the importance of separating the various moral concerns about

manipulation (e.g., perhaps the spouse example does pose a threat to autonomy, but is not morally

concerning all-things-considered for other reasons), and also about the importance of the context in which

the manipulation occurs, particularly the context of the relationship between the two parties.

p. 131

Thus, the reason why we �nd the spouse case generally morally unproblematic and the government case

potentially concerning is the di�erence in duties, obligations, and expectations that arise out of those two

relationships. It is entirely reasonable to think that the wife has an obligation to care for the well-being of

D ow

nloaded from https://academ

ic.oup.com /book/4870/chapter/147239951 by Florida State U

niversity user on 04 January 2024

her spouse, and that to some extent he has an obligation to care for his health for her bene�t. Moreover, we

expect this sort of non-argumentative in�uence from our spouses, and view it as occurring fairly routinely

in our interactions. It is less clear that the state has an obligation to care for the well-being of its citizens

in anything more than a negative sense of protecting them from harm, nor that the citizens have an

obligation to care for their health for the sake of the state.

23

Before moving on, I want to note a way in which the relationship or legitimacy factor ties the other factors

together. The extent to which it is morally problematic if autonomy is not protected or promoted, or good

ends are not promoted, depends in part on the relationship between the in�uenced and the in�uencer. For

example, a doctor may have an obligation to work to enhance a patient’s autonomous decision making,

whereas an agent of the state, or an advertiser, does not have that obligation. Thus, a case of non-

argumentative in�uence used by a doctor that threatens autonomy to some extent and fails to promote the

patient’s best interests is morally worse than a case of non-argumentative in�uence used by a state

institution or advertiser that threatens autonomy to some extent or fails to promote the subject’s best

interests. Moreover, the doctor who knows his or her patient well is more likely to be able to use non-

argumentative in�uence to produce good consequences for the subject of the in�uence since she knows the

patient’s goals and values, which is perhaps not the case with the state or an in�uencer far removed from

knowledge of the subject’s goals and interests.

p. 132

The Challenge of a Pluralistic Framework

One challenge to any pluralistic account of the ethical permissibility or impermissibility of non-

argumentative in�uence, such as the one I am suggesting, is how to deal with cases where an instance of

non-argumentative in�uence is problematic with respect to one normative feature (e.g., autonomy) but not

another (e.g., promotion of good ends and virtue over vice) and to come to an all-things-considered

judgment about ethical permissibility. This is a challenge that needs to be taken up, given that, as I have

argued, the moral status of the wide variety of practices encapsulated by the term “manipulation” depends,

and depends on a wide variety of factors. It is not always the case that in�uencing another non-

argumentatively by bypassing or countering his reasoning capacities is morally wrong, nor is it even always

the case that it infringes on autonomy or involves vices such as disrespect or arrogance. Moreover, even if

an instance of non-argumentative in�uence did, say, infringe on autonomy, we may still �nd it morally

permissible for other reasons, given that autonomy is not the only appeal that generates normative reasons.

So, we are left with a much more complex picture that requires us to evaluate particular instances based on

the moral framework and then come to an all-things-considered judgment about the ethical permissibility,

which will involve weighing some moral reasons against others.

24

I am not going to resolve this issue in this essay, but let me make some suggestions. There are several

avenues one can pursue in response to this challenge: (1) commit to one of the moral reasons (e.g.,

autonomy) in an absolutist sort of way or grant it “lexical priority” in a Rawlsian sense; (2) take a prima

facie approach inspired by W. D. Ross; or (3) take a moral particularist approach inspired by Jonathan

Dancy. I am going to set aside the absolutist approach, for it is the easy (yet most unattractive) response,

and instead focus on making an initial sketch of the Ross and Dancy approaches to the problem. In The Right

and the Good, Ross developed a theory of prima facie duties such that there is no master moral principle or

duty, but instead that there are a plurality of duties and moral reasons that we ought to consider in

determining what we ought to do, all things considered. Ross did hold, however, that some of these duties

are more important than others (e.g., the duties related to personal relations, such as non-male�cence,

�delity, reparation, and gratitude are more weighty than the duty to promote a maximum aggregate of

good). That might o�er us some guidance, but unfortunately Ross does not provide us with an algorithm

p. 133

25

26

27

28

D ow

nloaded from https://academ

ic.oup.com /book/4870/chapter/147239951 by Florida State U

niversity user on 04 January 2024

Notes

for coming to an all-things-considered determination of what we ought to do, but instead refers to the

notion of a “considered opinion” that “rests with perception.”29

Dancy’s moral particularism is not going to o�er us an algorithm either, but it does o�er an additional tool

that may be useful. Particularism allows for a feature to have one moral valence in one case and another in a

di�erent case. So, for example, impairment of autonomy might have more of a negative moral valence in

one case of non-argumentative in�uence than in another. For instance, in a case where we think that the

obligations and expectations regarding autonomy that arise out of the relationship between the in�uencer

and the in�uenced are strong (e.g., physician–patient), non-argumentative in�uence that impairs

autonomy may have more of a negative moral valence than in one where they are weak (e.g., advertiser–

consumer, or even wife–husband). If this is correct, then one of the places to start with the assessment of

the moral status of an instance of manipulation or non-argumentative in�uence is by assessing the

obligations and expectations arising out of the relationship between the in�uencer and the in�uenced,

considering them as one moves through the moral framework and assessing the extent to which the

in�uence threatens or promotes autonomy, virtue and vice, and good aims and consequences over bad ones.

p. 134

Conclusion

The central moral question with respect to any practice described as “manipulation” is whether it is

ethically permissible. Practices described as “manipulation” are better described as practices of “non-

argumentative in�uence,” since this description avoids any automatic and undefended assignments of

negative moral status. Non-argumentative in�uence is in�uence that either bypasses or counters a person’s

reasoning capabilities. The moral status of non-argumentative in�uence depends and depends on the extent

to which the instance of in�uence (1) threatens or promotes autonomy; (2) has good aims and virtuous

overtones or bad ones; and (3) ful�lls or fails to ful�ll duties, obligations, and expectations that arise out of

the relationship between the in�uencer and in�uenced. This framework is a pluralistic one, and presents

the usual challenge of pluralistic frameworks, which is “weighing” some moral reasons against others.

Moral particularism, with a focus on spelling out the duties and obligations that arise from the relationship

between the in�uencer and the in�uenced in a particular case, is a useful theoretical framework for helping

us to come to an all-things-considered judgment about the ethical permissibility of non-argumentative

in�uence in a particular context.

1 Ruth R. Faden and Tom L. Beauchamp, with Nancy M. P. King, A History and Theory of Informed Consent (New York: Oxford University Press, 1986), 261; Marcia Baron, “Manipulativeness,” Proceedings and Addresses of the American Philosophical

Association 77, no. 2 (November 2003): 29, 42 10.2307/3219740 ; Robert Noggle, “Manipulative Actions: A Conceptual and Moral Analysis,” American Philosophical Quarterly 33, no. 1 (January 1996): 43–55; Joel Rudinow, “Manipulation,” Ethics 88, no. 4 (July 1978): 338–47 10.1086/292086 ; Claudia Mills, “Politics and Manipulation,” Social Theory and

Practice 21, no. 1 (Spring 1995): 100, 108–10 10.5840/soctheorpract199521120 ; Judith Andre, “Power, Oppression and

Gender,” Social Theory and Practice 11, no. 1 (1985): 111 10.5840/soctheorpract19851116 . 2 J. S. Blumenthal-Barby, “Between Reason and Coercion: Ethically Permissible Influence in Health Care and Health Policy

Contexts,” Kennedy Institute of Ethics Journal 22, no. 4 (2012): 345–66. 3 Melissa Seymour Fahmy, “Love, Respect, and Interfering with Others,” Pacific Philosophical Quarterly 92, no. 2 (June 2011):

183. 4 Eric M. Cave, “Whatʼs Wrong with Motive Manipulation?” Ethical Theory and Moral Practice 10, no. 2 (December 2006): 138.

Cave reiterates this position in this volume, but here he allows for a “broad” notion of modest autonomy such that someone could consent to have his or her motives or concerns managed by another via non-argumentative influence. In

D ow

nloaded from https://academ

ic.oup.com /book/4870/chapter/147239951 by Florida State U

niversity user on 04 January 2024

this rare instance, non-argumentative influence (what Cave calls “manipulation”) would be compatible with autonomy; “Unsavory Seduction and Manipulation,” this volume, xxx–xxx).

5 James Stacey Taylor, Practical Autonomy and Bioethics (New York: Routledge, 2009), 7–8. 6 Taylor, Practical Autonomy and Bioethics, 41. 7 Taylor, Practical Autonomy and Bioethics, 4. 8 Sarah Buss, “Valuing Autonomy and Respecting Persons: Manipulation, Seduction, and the Basis of Moral Constraints,”

Ethics 115, no. 2 (January 2005): 212 10.1086/426304 . 9 Moti Gorin makes a similar point in this volume (“Towards a Theory of Interpersonal Manipulation,” xx–xx), arguing that

manipulation and autonomy are compatible on internalist accounts (since an agent could be manipulated to have her first- and second-order desires achieve coherence and hence to be autonomous) and on mixed externalist (i.e., attitudinal historical) accounts such as Christmanʼs (since it is an open question as to whether or not the agent authorizes the process by which her desires or behaviors were formed). Thus, Gorin, myself, Buss, Christman, and Frankfurt all hold a similar position on the compatibility of manipulation and autonomy—namely, that it may be the case that an agent is manipulated (or influenced via non-argumentative means by another agent) and still autonomous. John Christman, “The Historical Conception of Autonomy,” in The Politics of Persons: Individual Autonomy and Socio-Historical Selves (New York: Cambridge University Press, 2009).

10 Christman, “Historical Conception of Autonomy,” 11. 11 Harry Frankfurt, “Frankfurt-Style Compatibilism: Reply to John Martin Fischer,” in Countours of Agency: Essays on Themes

from Harry Frankfurt, edited by Sarah Buss and Lee Overton (Cambridge, MA: MIT Press, 2002), 28. 12 Cave also resists this equivalence in his essay in this volume (“Unsavory Seduction,” xxx), writing, “from a moral

standpoint, we can distinguish between nonrational motivational change or stasis resulting from either environment or non-negligent accident and such change or stasis that is planned, anticipated, or caused accidentally but negligently by another. If the above observations about modest motive autonomy are correct, then cases of the latter evince a morally reproachable form of disrespect.”

13 Colin McGinn, Mindfucking: A Critique of Mental Manipulation (Durham, UK: Acumen, 2008), 46–47. 14 Baron, “Manipulativeness,” 50. 15 Baron, “Manipulativeness”; Fahmy, “Love, Respect, and Interfering with Others,” 184. 16 Aristotle, The Rhetoric and Poetics of Aristotle (New York; McGraw-Hill, 1984), 22. 17 Fahmy, “Love, Respect, and Interfering with Others,” 179. 18 Noggle, “Manipulative Actions,” 52–53. 19 Baron, Wood, and Manne essays in this volume. 20 Patricia Greenspan, “The Problem with Manipulation,” American Philosophical Quarterly 40, no. 2 (April 2003): 160; McGinn,

Mindfucking, 27–28; Stanley I. Benn, A Theory of Freedom (Cambridge: Cambridge University Press, 1988),

134 10.1017/CBO9780511609114 . 21 Baron, “Manipulativeness,” 48. 22 John Wryobeck and Yiwei Chen, “Using Priming Techniques to Facilitate Health Behaviours,” Clinical Psychologist 7, no. 2

(October 2003): 105–108 10.1080/13284200410001707553 . 23 Marcia Baron makes a similar point about manipulation in this volume (“The Mens Rea and Moral Status of Manipulation,”

xxx), arguing that one circumstance in which it might be justified is when there is a normative expectation of it. 24 Incidentally, most other authors in this volume end up relying on a pluralistic account as well (e.g., holding that even if

manipulation violates autonomy or respect [for reasons or for agents], it may still be morally justified on other grounds). Cave is explicit about this pluralism, noting that autonomy is not the only or overriding good (“Unsavory Seduction,” xxx), as is Baron, noting that manipulation may be justified if it is the lesser of two di�erent moral evils (“The Mens Rea,” xxx).

25 John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971). 26 W. D. Ross, The Right and The Good (New York: Oxford University Press, 1930). 27 Jonathan Dancy, Ethics Without Principles (New York: Oxford University Press, 2006). 28 Ross, The Right and The Good, 19–30. 29 Ross, The Right and The Good 19, 30–33, 42.

D ow

nloaded from https://academ

ic.oup.com /book/4870/chapter/147239951 by Florida State U

niversity user on 04 January 2024