Then again, what is manipulation? A broader view of a much-maligned concept

ABSTRACT

We influence each other constantly and in diverse ways. At times ethically, as when we convince others via arguments founded in good reason. At times problematically, as when we coerce others to act in a certain way. Other forms of influence, such as manipulation, lie in between these poles, as when we influence others not primarily rationally but also not coercing someone, but indirectly by modulating their affective states. Manipulation is usually associated as a deceptive, harmful and sneaky form of influence that is morally problematic. In this paper, I want to distinguish between objectionable and unobjectionable attempts of manipulation and suggest a new, integrative account of manipulation that offers a broader view of a much-maligned concept.

KEYWORDS:

• Manipulation
• affects
• rationality
• deception
• coercion
• influence

1. Introduction: should we name a non-rational and non-coercive attempt at influencing an agent ‘manipulation’?

We influence each other constantly and in many ways. On the one hand, there are ethically unproblematic forms of influence, e.g. if we want to convince someone of something in a rational, argumentative manner founded in good reason. On the other hand, there are those we often identify as problematic as they interfere with our capacity for acting freely: forms of influence that coerce us to act in a certain way. But it seems that there are also forms of influence that lie in between these poles in a continuum of influence, not primarily rationally influencing us but also not coercing us. They make use of our inner, often unconscious workings, using peripheral routes of decision-making to influence by means of modulating our affective states, more precisely: our feelings, emotions and moods.¹

This is strongly connected to a phenomenon we call ‘manipulation’ that carries with it a flickering negative connotation. We encounter the term ‘manipulation’ in many areas such as advertising, politics and our personal relationships. While manipulation has traditionally been viewed quite negatively, typically as a method of deceitful coercion, it is in fact a more complex affair. In this paper, I provide a broader and – I hope – more differentiated view of the phenomenon that accounts for more innocuous and even benevolent iterations. I will offer a new understanding, called the Pleasurable-Ends-Model of Manipulation (PEM). Accordingly, manipulation will be understood as a form of influence where a manipulator tries to modulate the affective states of the manipulated, suggesting some end (an action, a product etc.) that is perceived as pleasurable/unpleasurable, but whereby the manipulated remains, at least in a minimal sense, free to choose or not to choose this end. Ultimately, I will argue that manipulative strategies that target the affective register are not necessarily at war with rationality and autonomy, but rather are a constitutive part of a complex decision-making process that draws both upon affective states and reason, and not always upon one at the expense of the other. While manipulation may restrict and modify rational freedom, it does not, we will see, necessarily render the subject unfree. I will try to conceptualize manipulation in accordance with Aristotle’s understanding of rhetoric, as a neutral tool employing problematic and unproblematic tactics to be used by virtuous or depraved persons, for good and bad purposes, since basically almost everything, except virtue or the good will itself, can be misused. The same is the case, in my view, for manipulation. Of course, this needs explaining as it will seem to many counter-intuitive.

Conceptualizing manipulation is no easy task, not just because there are many intuitions involved. The discussion about how manipulation as a form of influence can be understood (and how it should not be) is at times difficult because there are – even though this is a rather small field of study – many varying accounts of manipulation, some more elaborated, but most of them not so much. In light of the general confusion, Theun A. van Dijk criticized the lack of a thorough definition of manipulation over 20 years ago: ‘although frequently used, the notion of manipulation has […] never been made explicit in a theory’ (van Dijk 1998, 340). Joel Rudinow, the author of an example I will get to in a bit, laments this already in 1978 (Rudinow 1978, 338).² This lack of a clear definition is especially striking given the ubiquitous presence of manipulation as a form of influence. So, it is a particularly difficult and nonetheless important thing to discuss manipulation’s mechanisms and its implementations since it just seems to be unclear what it consists of as a strategy and which specific tactics it includes. The ambiguity of the existing concepts results in making the word ‘manipulation’ handily usable as denouncing an attempt to influence – without discussing it further. In this understanding, selfish manipulators intentionally operate in a sly and underhanded manner, with deceit, leading a manipulated agent to negative, harmful ends – just as Iago in Shakespeare’s Othello does. Furthermore, manipulation bypasses and even subverts the manipulated’s rationality – something that all of the examples about manipulation seem to have in common to some degree – thus harming an agent’s autonomy, treating her in an undignified manner (this, in turn, can be discussed regarding the specific cases) (Fischer 2017, 16).

This is far from what ‘manipulation’, which stems from the Latin manipulus and meant ‘a handful’ (of soldiers in a legion), once entailed. After being used as meaning ‘holding someone by the hand’ (in an even helpful manner!), it became a technical term meaning that something ‘is handled skillfully’ (e.g. a machine; we are getting closer to our contemporary meaning). It is in the proximity of behaviorism and psychoanalysis that the term finds its way into scientific contexts as a result-oriented handling of a person. After an eventful history of dictatorships using propaganda to form public opinion with fear, anger, mistrust in others, and admiration for evil authoritative figures, it is understandable that the above-mentioned associations made their way into our ordinary language use of the term ‘manipulation’. The neo-Marxist tradition and a general shock about the past effectively earmarked it as an accusation that a devilish manipulator is influencing an agent in a reprehensible manner by deceiving and subverting rationality, which led to ‘manipulation’ having (negative) moral connotations (Fischer 2017, 36–40). Why, then, still speak of ‘manipulation’?

There are some reasons to keep the term but also to refine its understanding. For one, suggestions to not speak of ‘manipulation’ anymore raise the question whether it is possible to meaningfully speak about the phenomenon without this term. Readers would just not know anymore what we are talking about. There seems at any rate to be no other everyday-language term that pinpoints the modulation of our affective states. ‘Influence’ itself is a broad overarching category and too vague. ‘Affective influence’ or ‘non-rational influence’ might be neutral candidates which are less pejoratively burdened than ‘manipulation’ but certainly not commonly used. ‘Manipulation’ with its inclusion of the hand (like turning a dial) and its result-orientation still seems a nice fit. The negative connotation, however, might have more to do with the discussion overstressing a dichotomous concept of rationality and affective states with the result that challenging our rationality, ‘contaminating’ it with affectivity is seen as threatening our autonomy and dignity. Since we are all finite, boundedly rational beings, it nonetheless seems that we could admit that manipulation as a form of influence that aims at our affective states can be understood as a normal mode of communication, being used constantly to varying degrees. The term ‘manipulation’ entails this well as it marks an influence that goes beyond our rational radar. In fact, by adding some differentiations, the term does not sail too far from its everyday use, and becomes more precise. Interestingly, in a recent survey conducted by a colleague from marketing and myself, a representative group characterized ‘manipulation’ as neither necessarily morally problematic nor necessarily deceptive and harmful. It was generally seen as a form of influence using our affective states where, because it uses this avenue, we should be attentive as it might become problematic at some point (Feurer and Fischer 2022). So, the question whether the negative moral valence is essential to manipulation, or if giving a neutral account of the phenomenon is plausible is all the more relevant, not just because it is rational to differentiate as clearly and precisely as possible, but also because empirical evidence seems to support this. Consequently, manipulation might seem like a prima facie wrong attempt to influence with certain questionable tactics, but if we look closer, this is not at all as clear as it is assumed. Instead of assuming that manipulation is necessarily morally problematic, deceptive, harmful and sneaky, I want to distinguish between objectionable and unobjectionable attempts of manipulation – the following examples will illustrate some but certainly not all of them (there are many more to find).

So, let us consider some examples of influencing an agent by primarily non-rational but also non-coercive means that I want to suggest as being manipulative at their core, pointing out that the attempt to modulate our affective states by rendering an end as pleasurable/unpleasurable seems to be at the heart of the phenomenon we call ‘manipulation’, even though there are several other features like sly and underhanded tactics, deception, harmful, negative ends and selfish manipulators that over time became associated with it:

1. First, consider Iago as the par excellence manipulator from William Shakespeare’s Othello: Iago famously has a plan that will lead to a devastating end: he is framing Cassio and Desdemona as lovers to make Othello jealous – ‘[…] as tenderly led by th’ nose/As asses are’ (I.iii). He aims at kindling this fire of jealously, an affective state that makes us especially vulnerable and where ‘Trifles light as air/Are to the jealous confirmations strong/As proofs of holy writ’ (III.iii.). Othello soon becomes upset and moody. Iago continues to further his goal of removing both Cassio and Othello by hinting again and again (with words but also with objects – such as the famous handkerchief) that Cassio and Desdemona are involved in an affair, strategically using the emotions of jealousy and anger as well as the depressed mood Othello gets into for Iago’s ‘sport and profit’ (I.iii). Desdemona’s attempts to reinstate Cassio as lieutenant add to Othello’s almost immediate conviction that his wife is unfaithful. Through Iago’s complex manipulative strategy and his various tactics, Othello becomes consumed by his jealousy, anger and mistrust. Emilia observes righty: ‘But jealous for they’re jealous. It is a monster/Begot upon itself, born in itself' (III.iv). Othello kills Desdemona in a fit. In the end, Iago has sewn chaos and sees his attempt at extensive revenge fulfilled.

2. Secondly, consider this legendary example from German history: the attempt to make the potato a popular food in Europe turned out to be a difficult thing, even when it was supposed to be the solution to a severe hunger crisis in Prussia in 1746. The farmers did not embrace the new product, because inedible types circulated and nobody really knew how to cook them correctly. After unsuccessfully trying to command the growing of potatoes, the Prussian government used different strategies to get their people to accept the new vegetable. Priests had to preach the value of the potato, and Prussian King Frederick the Great tried to frame potatoes as something delicate and valuable, e.g. by having soldiers guard a field where potatoes were grown. In France, Antoine Augustin Parmentin tried to solve the same problem against the background of several hunger crises. He supposedly even organized several public dinners where famous individuals ate in front of huge crowds, praising the potato – with a success that made the potato one of the most popular foods. Average people became curious and a craving for the ‘fine’ product developed relatively quickly and significantly helped confine hunger.

3. Thirdly, consider a snippet of Mark Twain’s Tom Sawyer where Tom has to whitewash a fence and heavily dreads his duty. His (rather lazy) attempts to rationally convince Jim to take over for him fail quickly. He realizes that he can’t buy himself out and he cannot coerce someone into doing it either: ‘At this dark and hopeless moment an inspiration burst upon him! Nothing less than a great, magnificent inspiration!’ Tom instantly deploys that inspiration when Ben comes up to him asking what he’s doing: ‘No answer. Tom surveyed his last touch with the eye of an artist; then he gave his brush another gentle sweep and surveyed the result, as before. […]’ Ben tries again: ‘‘Hello, old chap, you got to work, hey?’ Tom wheeled suddenly and said: ‘Why it’s you, Ben! I warn’t noticing.’ ‘Say–I’m going in a-swimming, I am. Don’t you wish you could? But of course you’d druther work–wouldn’t you? […] Tom contemplated the boy a bit, and said: “What do you call work?” “Why, aint that work?” Tom resumed his whitewashing and answered carelessly: “Well, maybe it is, and maybe it ain’t. All I know, is, it suits Tom Sawyer.” “Oh come, now, you don’t mean to let on that you like it?” The brush continued to move. “Like it? Well I don’t see why I oughtn’t to like it. Does a boy get a chance to whitewash a fence every day?” That put the thing in a new light. Ben stopped nibbling his apple. […] “Say, Tom, let me whitewash a little”’. The rest is known. Tom does not have to do a stroke anymore and even makes a profit off letting others whitewash. Since rational arguments did not work for him and coercion was not an option, Tom makes the whitewashing appear as fun, artistic and even a privilege – instead of arduous and monotonous work.

4. Fourthly, consider this scenario: you call your grandmother and she, after letting you know that you did not call for a long time, tells you that she is so lonely since her husband died. At a certain point, your nana asks in a pitiful and slightly accusatory tone when you will finally pay her a visit after such a long time. This is how we can interpret this: she is trying to induce guilt (maybe not even consciously; this is just how she is able to utter her needs), not by saying that this or that is wrong and you should do otherwise (hence arguing), but by suggesting something using a certain tone and a choice of non-causal words to make you acknowledge that her life is not easy, to make you believe that you could at least do a little bit to ease that pain, and thus moving you towards visiting her – again without primarily convincing you rationally. You might see what she is doing, but it works anyhow. You feel guilty and look at your calendar as to when you can carve out the time.

5. Fifthly, let us consider an example by Joel Rudinow, who wrote one of the first papers on manipulation as a specific strategy to influence an agent: ‘Jones complains of being the object of regular manipulative attempts by his wife. He describes a typical instance of this. He is making ready to leave the house for his weekly poker game, of which he is very fond. It has been an uncommonly busy period for Jones; he has not spent one evening in the past two weeks at home with his wife. His wife now appears, clad in a see-through nightie, poses seductively, begins nibbling at his ear lobe and playing with the buttons on his shirt. Jones protests that he does not want to be tempted just now – he will end up missing his poker game. Furthermore, Jones suggests that his wife is really not at all interested in sex. Though he has arrived home late at night for the past two weeks, the couple’s sexual frequency has been higher than at any other time in the couple’s history. He accuses her of attempting to manipulate him so that he will remain at home with her for the evening. His wife admits to this immediately, adding that she knows as well as Jones how frequently they have had sex of late and that he knows as well as she does that she knows this as well as he does' (Rudinow 1978, 341). Here we can find the classical complaint against Jones’ wife because she tries to manipulate him. She thinks that trying to change his actions works effectively by modulating his affective states, in other words: inducing lust in him (even though he is now appalled, so she might have overdone her attempt) instead of arguing with him. Yet both are well aware of what is happening and everything is out in the open.

6. Last but not least, let us consider a rather obvious scenario of non-rational but also non-coercive influence. Punishment and gratification seem like common tools in our interactions. We can understand punishment as something that is done after a given deliberate action that lowers the chance of that action taking place in the future. It is meant to discourage a certain behavior and appears in various forms in relationships. Imagine a couple where George, upon his return, wants to spend time with his partner Cynthia after feeling alone during some time on a work-trip. But Cynthia has promised a friend to be there for her on that exact evening of George’s return, since Cynthia’s friend had a difficult and problematic negotiation with her mother who disapproves of her partner. Even though Cynthia informed George beforehand that she might have to stay with that friend for some time, he is disappointed. Rationally solving the problem (Cynthia presents reasons that this is an emergency, that she announced it beforehand, that she will only be gone for a few hours; George says that he has been alone for so long and just cannot take it any more) is not successful. Consequently, George freezes her out and shuts down their communication that evening. This can be considered as the form of punishment we usually call love-deprivation. George tries to sew doubt as to whether Cynthia did the right thing (which works well as she is doubting herself constantly). He also induces guilt and displays his hurt for her to deal with. On the opposite side, gratification encourages certain behavior, e.g. making it more likely that someone will eat healthy or learn something with some relish. Neither punishment nor gratification are means to rationally convince an agent or to coerce her to act in a certain manner.

The potato-legend, Tom Sawyer’s escape from whitewashing the fence, George’s love deprivation, nana’s guilt trip, and Jones’ wife’s seduction may at first sight not seem like paradigm examples of manipulation, whereas Iago’s handling of Othello is usually considered the classic example. Iago’s attempt at influencing Othello brings with it selfish, even devilish motives, heavy deception and underhandedness, so that Othello is unaware of what is happening to him, basically without much of a chance to act freely anymore and facing a gruesome end (even though he retains a modicum of agency – otherwise the play would not be a tragedy³). It has ‘morally problematic’ written all over it. And this is what manipulation is usually thought to be, and in our everyday-use of the term is frequently accused of. The other examples seem to show a different ‘degree of problematic’.

All of these examples have one thing in common, though: there is an agent (to a certain degree) intentionally trying to change how someone acts neither by rational argument nor coercion but via a primary attempt to modulate an affective state by presenting an end as pleasurable/unpleasurable. Iago creates the very unpleasurable fantasy of Desdemona being unfaithful; nana evokes the image of her being lonely (and a little hurt); and George the prospect of Cynthia having to deal with a grumpy husband and maybe even with her own insecurity and guilt. Tom Sawyer and Jones’ wife, though perhaps selfish (but in the wife’s case without any underhandedness or deception), unlike Fredrick the Great/Parmentin, try to modulate the affective states of their targets by presenting whitewashing, eating potatoes, or staying home as very pleasurable. It immediately becomes obvious that we can find a varying degree of intention (and effort), such that some attempts are not 100% manipulative but rather exhibit a manipulative quality while mixed with other forms of influence such as trying to rationally persuade (e.g. Tom mixes rational persuasion and manipulation). It is also obvious that different tactics are employed to support manipulating someone such as underhandedness, deception and different (maybe even harmful) outcomes for the manipulated. Othello’s fate presents a catastrophic result, whereas in the other cases the consequences are not nearly as devastating and maybe even harmless. On the whole, our attempts to influence each other are rather often combined which is why I choose the image of a continuum, where no part can be distinguished from neighboring parts in an absolutely clear-cut manner, except by rather arbitrary division. To get a clear sense of the different forms of influence, it is nonetheless rational to apply ideal differentiations while keeping in mind that real life is often messier than philosophical categories. It is not just difficult to conceptualize how we influence each other in a tidy, clear-cut manner, as the world is a ‘great blooming, buzzing confusion’, as William James once put it (1890, 488), but also because human decision-making is something of a rather complex black box-shaped by concrete situations, specific contexts, habits and individual character traits (Fischer 2017, 63, 85). Operant conditioning with its use of punishment and gratification is perhaps the most clear-cut example. But there are many more ways to use our affective states in the realm of influence than conditioning, as the examples above illustrate.

The encumbered everyday-use of the term ‘manipulation’ with its characteristics may not necessarily hold up as necessary and sufficient conditions of manipulation. But we should not just succumb to our intuitions or leave it to custom. I propose that it is rational to take a step back to better understand the phenomenon by looking at manipulation on a descriptive level. This might shake up our intuitions and questions an understanding clouded by rather extreme and well-known instances of manipulation (besides Othello, Orwell’s Nineteen Eighty-Four might come to mind). I want to consider less extreme and even benevolent cases, where the manipulated is not harmed and even where her/his welfare is considered. The negative valence might in the end not even lie in the manipulation itself, but rather depends on how exactly a manipulative attempt is played out, and if and why it includes underhandedness, deception and self-serving manipulators in a specific case.

2. Towards a thorough definition

We can start off our attempt to come up with a thorough definition of manipulation by generally considering it as a broad and ubiquitous form of influence of human agency which in its many forms aims at shaping our affective states and consequently also our thinking and acting. By and large, manipulation is one of various ways to influence our capacities as acting agents. On a coarse map of influences the categories rational persuasion and coercion are well explored. But manipulation seems to be a form of influence somewhere in between these two poles (Beauchamp 2001, 477; Fischer 2017, 53). Rational persuasion and coercion can be understood as poles because they mark the outer rims in a continuum of how we are influenced in making a decision. These poles are to be understood more like brightness and darkness and less as a clear-cut dichotomy. Thus, there are many shades in between making an influence more or less rational. In the case of coercion more and more alternatives are drowned out and thus coercion is born. In the case of coercion (which can be executed in many different ways) an agent is not able to decide freely in an extensive sense or sometimes not even rationally because she is presented with no (at least preferable) alternatives and possible harm, if she does not act like the coercer wants him to act. Tom, George, nana, Frederick the Great and Jones’ wife do not do this, whereas Iago with his extensive strategy makes it almost impossible for Othello to act differently. Almost. The case of rational persuasion presents us with the opposite: Here, an agent is enabled to decide freely on the grounds of good reasons which are presented so she can make her deliberation based on all the available information in line with her rational capacities.⁴ It seems like none of the above examples show a clear case of rational persuasion, only attempts to argue to a certain degree. Tom is applying some argument, George tries to employ a reason to make Cynthia stay (but it does not work so he changes his tactics), nana offers implied arguments (while primarily aiming at your affectivity), Jone’s wife and Iago do not even care about arguments and Frederick the Great fails to make a mark with his good reasons. Coercion marks the limit of a free decision-process, whereas rational persuasion can be considered as the maximum and an ideal decision-making process because we are neither threatened nor is our deliberation contaminated by any other factors that are possibly difficult to control like affective states. All the agents in our examples find their affective states modulated as jealousy, guilt, desire, insecurity and curiosity are induced, but are at least in a minimal sense still free to act. Manipulation thus seems to be something that does not in a full sense coerce or rationally persuade an agent. Our affectivity plays a notoriously ambivalent role in the debate about the nature of free decision-making and acting, ranging from being an important part of it to being a grave threat – or just irrelevant to it. This might have a lot to do with thinking of our rationality and our affectivity in the sense of a strict dichotomy, which does not seem plausible. Our affectivity plays an important role for a reasonable decision-making process, as it shapes a meaningful perception of the world, helps us to judge, identify values and be motivated. It also shapes our thoughts (‘Should I really visit nana soon?’, ‘Desdemona is betraying me in the most gruesome way!’), and our thoughts are in a constant interplay with how we feel (the picture of Desdemona betraying Othello might deepen his pain and jealousy; similarly Cynthia might worry a lot and play out scenarios in her head – emotions like jealousy and fear trigger especially creative cognitions).⁵ Nonetheless, a decision that is worthy to be called ‘free’ also means being able to lift ourselves from our affective states and certain thought processes (e.g. dysfunctional patterns like catastrophizing) to ultimately gain an integral decision as the foundation of our action.

Let us briefly look at the general discussion on manipulation and its varying definitions before establishing a different, integrative account of manipulation, which is inspired by earlier philosophical definitions, action theory and psychological research, but also tries to avoid some of the stated problems other conceptions of manipulation and our everyday-use of the term face. In consequence, manipulation will be understood as intentionally modulating the affective states of an agent by presenting a certain end (an action, a product etc.) as pleasurable/unpleasurable, which entails the suggestion to choose or not to choose this end, but where the manipulated remains, at least in a minimal sense, free to act in this manner or not. The active (often not fully conscious) curation of the affective attraction of certain ends or their realization makes some options more (or even extremely, whereas others not at all) appealing and, consequently, more likely to be chosen (Fischer and Illies 2018, 27). As our affective states (feelings/emotions/moods) are actively modulated, the evaluation of a certain end changes, resulting in often complex affective experiences which eventually boil down to a desire or an aversion, a yes or no, that motivates us to act in a manner according to the manipulator’s goal (or not). Even Othello’s jealousy, which creates fierce affective turmoil, channels into certain (in this case dysfunctional) agency. This might seem like a rather atypical definition of manipulation, as it leads away from our everyday understanding at least a bit. Hence, let us walk through the individual steps that lead to this proposed definition.

3. The different concepts of manipulation

3.1. Manipulation as underhanded and deceptive

Underhandedness and deception go well together as deceiving someone usually happens under an awareness threshold – otherwise it would hardly work. They are, however, two separate features, usually assigned to manipulation, even though differentiating them is not always easy. Deception is an effort to unknowingly create an error in an individual’s perception of reality; underhandedness can be seen as a supportive tactic used to (more or less skillfully) hide something and prevent it from being known. Thus, I will be looking at both of them briefly.

Manipulation, as we have seen, has the reputation of being something underhanded and sly. It supposedly happens in disguise, under the awareness threshold of the manipulated, who is tricked to act in a certain manner without the knowledge of being influenced. Thus, manipulation is often characterized as an ‘opaque instrument of power […] of which recipients are not or barely aware, or of which they cannot easily control the consequences’ (van Dijk 1998, 275; see also Goodin 1980, 9; rather similar views are presented by e.g. Ware 1981, 165; Baron 2003, 37). But this does not help much in characterizing manipulation accurately, not just because almost anything hidden and tricky could thus be called ‘manipulation’ (which does happen) – like cheating in a game, lying, or even performing magic. But also because there are many cases of manipulation in which underhandedness does not play a vital part. Nana’s guilt trip, George’s love deprivation or Jones’ wife can count as examples of that. They do not even employ truly wily tactics. Hence, there are authors that object to underhandedness as a necessary condition of manipulation and argue that ‘[w]e can be lucidly aware that we’re being manipulated [e.g] into feeling guilt and act on it’ (Barnhill 2014, 59) and sometimes there is not even an intention to hide a manipulation such as in the case of Jones’ wife. Romantic relationships on the whole seem like a realm for obvious and open manipulation that we engage with happily. We can also think of consciously manipulating ourselves as a form of self-therapy to make us feel better or gain a healthier attitude. Consequently, underhandedness does not qualify as a necessary and sufficient condition of manipulation. However, it can be an amplifying condition, which might help to manipulate more effectively. Tom in his rather tactical way, does not disclose his motivation to get someone else to whitewash the fence. By saying ‘I don’t see why I oughtn’t to like it’, Tom entices Ben to reflect on all the positives that allegedly come with painting a fence, which, he suggests, is not drudgery but rewarding artwork that cannot be done every day. This seems to be enough for Ben to give in, and he is all the more ready to take over when Tom starts to hesitate in regard to letting him take over.

Tom does not convey his motives and plays with the suggested pleasantness that whitewashing a fence supposedly entails. Iago a fortiori does not convey his evil plans and actively works to alter Othello’s perception. Manipulation is usually seen to be at least closely related to deception (sometimes both words are even used interchangeably).⁶ Thus, Thomas Scanlon for example defines manipulation as a means of inducing false beliefs and expectations (1998, 298) which boils down to what we usually call ‘deception’. Robert Noggle (1996) aims in a similar direction but shifts the perspective a little when he conceptualizes manipulation as an attempt to ‘lead someone astray from certain paths toward certain ideals’ (44, a rather similar view is to be found in Hanna 2015). Accordingly, manipulation is seen as a tool to lure someone away from certain ideals (even though it is not completely clear what kind of ideals these are), which is supposedly not in the self-interest of the lured. Therefore, every form of misleading (like lies or the selective presentation of information) can be seen as efficient tools for manipulating someone. Claudia Mills (and in a closely related manner Gorin 2014a, 2014b) finds other words to describe manipulation as essentially deceptive:

[…] manipulation in some way purports to be offering good reasons, when it fact it does not. A manipulator tries to change another’s beliefs and desires by offering her bad reasons, disguised as good, or faulty arguments, disguised as sound–where the manipulator himself knows these to be bad reasons and faulty arguments. (1995, 100)

Manipulation consequently not only induces false beliefs (e.g. that something is pleasurable for you) or lures you away from living up to certain ideals (e.g. that you wanted to refrain from buying too many things), it also deceives because it does not argue with good reasons, it only tries to let bad or false reasons seem acceptable (we could say: like Tom and Iago do), deluding the manipulated into a certain belief (like: whitewashing a fence is fun and a rare opportunity, or Desdemona is unfaithful). Here, rationality also plays a role because there are reasons (even though they are ‘false’ ones), among other things, at play. But there is a danger to characterizing manipulation simply as providing selective or false information. Understanding manipulation in this way reduces our chances to identify other types of manipulation. Of course, manipulation often seems to go well together with deception, which can also be qualified as a potentially amplifying condition, just as we can see with fake news: here the polarizing presentation of selected information usually aims at stirring up affectivity. It is possible to use lies, audio-visual illusions, or just present half of the facts in a greater scheme of manipulating an individual. Yet it is still possible to distinguish these two aspects: in regard to manipulation it can be based on a lie (‘I love whitewashing the fence!’, ‘Desdemona is unfaithful.’), or a lie can be used for amplifying a manipulation. However, there does not necessarily have to be deceit at play in order to manipulate someone, as all the propositions at play can be facts, but be used in a manipulative manner. Consider again Jones’ wife or nana, where nothing is hidden and the propositions are correct: Jones’ wife wants him to stay home, and she will openly use seductive tactics for that; nana is lonely and thinks you stayed away too long. It is the use of sexual desire in Jones’ case and nana’s arrangement of arguments reported in a pitiful tone that manipulate – in a different manner and to a different degree than Tom and Iago (which also surely differ in degrees and tactics – Ben is free to reject Tom’s resemanticization and the attached urge whereas Othello is less free) but aiming at the same thing: grabbing their targets by their affectivity. Thus, an appropriate account of manipulation should consider both cases: those where underhandedness and deception can be involved and those where they are not.

3.2. Manipulation as serving negative ends

In most accounts of manipulation, the term is also associated with the aim of fulfilling only the self-serving goals of a manipulator with potential harm for the manipulated. In other words: some agents try to achieve their selfish (and for the manipulated: negative) ends by manipulating others – whether it is someone guilting you into doing something for them or a company wanting you to buy their product so they can make a profit. In both cases there is someone with a tangible interest intentionally urging you to fulfill it, making you their ‘puppet on a string’ – the widespread paradigmatic (also negatively connotated) picture for illustrating manipulation (Wilkinson 2013, 342). This seems to be true for all of our examples to a different degree and, even though it might happen unconsciously at times (like in nana’s guilt-trip case), something that every manipulation entails. But even if there is a manipulator actively aiming at guiding you to a certain action, it seems a quite different question whether this is necessarily harmful for the manipulated. The neo-Marxist philosophers from the Frankfurt School provided this argument in regard to capitalism: various agents in modern societies with their amalgamation of state and capitalism use manipulation to prevent individuals from protesting against the unjust consequences of capitalism. This is done by constantly modulating their desire to consume and own things, e.g. by advertisements (Marcuse 1969, 13). Green and Pawlak (1983) point out another aspect which is already implied in Marcuse’s argument: manipulation is used by someone who is more skilled (or we could say: has more resources) than the target in order to exploit the weak and undermine their welfare (35; see also Sunstein 2015, 433).⁷

There is no doubt that there are many cases where manipulation involves willfully harmful or selfish behavior (in this manner Marcia Baron 2003 pinpoints ‘manipulativeness’, as a form of arrogance and vice) and serves negative ends. Iago is the example par excellence for this; Tom, George and nana might also be seen as selfish, but one can debate to what extent the consequences are actually harmful or render the target unfree. Frederick the Great/Parmentin on the other hand aim at ending starvation. Their cases illustrate the fact that it is reaching too far to assert that every kind of manipulation is harmful and only serves negative ends of a self-serving manipulator. It is also wrong to assume that every manipulator might have a corrupt character. The Tom Sawyer episode elucidates this: Tom is not a corrupt character, even though he is definitely a little rascal. Neither are George, nana and Jones’ wife (as far as we know), whereas Iago definitely is. The results of the manipulative attempts are also very different. It seems that the same mechanism that can lead people to bad deeds can also lead to good and even rather pleasant deeds. In some cases, think again of the potato-legend, manipulation might even substantially benefit its targets and help them to live healthy lives. At the same time, there seems to be no necessary benefit for the manipulator. Also think of romantic partners where one manipulatively furthers the welfare of the other without necessarily gaining an advantage for themselves. In this sense, manipulation is not necessarily a zero-sum game.

Hence, a more differentiated account is able to include both sides: the possible egoistic manipulator with his selfish and possibly negative ends (which also points us to its possible ethical illegitimacy!), and the opposite, which is often called ‘paternalistic’ influence, where a manipulator wants to nudge us to a supposedly laudable action mainly without rational persuasion or coercion. We just saw that underhandedness, deception, selfish manipulators, negative ends and harm need not be conditions of manipulation. However, most criticism regarding manipulation actually comes from another angle: the accusation that manipulation undermines our rationality and thus our autonomy.

3.3. Manipulation as bypassing and subverting our rationality

It seems to be commonplace that manipulation is ‘a kind of influence that bypasses or subverts the target’s rational capacities' (Coons and Weber 2014, 11; Fischer 2017, 41). That manipulation at least bypasses rationality to a certain extent seems to be plausible because it is something other than rational persuasion. It primarily seems to use our peripheral routes of decision-making, in other words: our affective states to get to us. Tom does this by playing with a set of ambiguities (‘Maybe it is, maybe it ain’t.'), nana presents us a picture to draw our own conclusions from on the grounds of a guilt-trip, George punishes Cynthia with love-deprivation, Frederick the Great/Parmentin cook up several events to raise the reputation of potatoes, Iago ensures that Othello becomes crazy jealous. In all of these cases, rationality is to a certain extent bypassed. The stronger case of subverting our rationality seems to be something different, however, and it may be present in Othello and also George Orwell’s 1984, where rationality seems to be destroyed or at least heavily constrained by Big Brother’s logic. What is safe to say is this: manipulation can be qualified as a form of influence that does not primarily involve the rational capacities of an agent. However, the assumption of many earlier accounts – which we will now take a brief look at – is that our affectivity and rationality are adversaries and inversely related. As stated above, I am arguing for a more mutual relationship.

Noggle for example suggests that manipulation is an act where the manipulator controls someone by adjusting their ‘psychological levers’ via (a) deception in order to change beliefs, (b) conditioning or changing situational factors in order to change desires (like Richard Thaler and Cass Sunstein suggest for nudging), or (c) the implementation of guilt or other emotions to change the manipulated’s feelings (43). Although it is not completely clear why only conditioning and changing situational factors have an impact on our desires (see Barnhill 2014 and Fischer and Illies 2018 for a critique of Noggle’s account), Noggle’s concept points in an important direction. Manipulation makes an individual fall short of ideals – though perhaps false ideals to begin with – for how to gain beliefs, desires and feelings and can thus be called a non-ideal response because of the use of psychological levers and not rational deliberation. A good influence, e.g. according to Gorin (2014b), cannot be achieved by manipulation, but is grounded in good reasoning and moves the addressee by virtue of these reasons (96). However, manipulation disconnects this link by making the manipulated act for the wrong reason – or even no reason at all. The Tom Sawyer-example shows this well. Tom succeeds in getting someone else to take over, who did not primarily deliberate rationally. (There might have been a rationalization after a feeling pulled Ben to whitewashing.) Tom also does not provide ‘good’ reasons, but rather creates an affective string to pull Ben. Against the notion of a normative ideal of decision-making, Tom generates a non-ideal response. The view that manipulation severs our ability to deliberate rationally is still very popular and much more can be said about this, especially about idealized views of rational ideals and the lack of an adequate theory of our affective states (instead of a simple binary view) which potentially leads to the caricatured view of manipulation and makes us oversee our own use of affective strategies to influence others.

In consequence, these accounts usually have a descriptive side (the manipulated does not ultimately act on the basis of good reasons because his rationality has been somehow bypassed or even subverted), but there is also a built-in normative evaluation. The ethical claim here is that manipulation does not ‘sufficiently engage or appeal to [agents’] capacities for reflective and deliberative choice' (Sunstein 2015, 443; my emphasis). This conclusion is drawn against the backdrop of an ideal of ‘sufficiency’. Joseph Raz even writes that manipulation ‘perverts the way that [a] person reaches decisions, forms preferences or adopts goals’ (Raz 1988, 377; my emphasis) and marks manipulation explicitly as unethical, not only by using a pejorative adjective, but also by contrasting it, again, with a certain ideal of normality.⁸

Indeed, manipulation is not the same as a rational argument and seems to not primarily work with our rationality. Thus, our autonomy (and maybe even our dignity) could be at stake. But at the same time, it seems to be an over-statement to say (a) that every manipulation subverts rationality. Gorin (2014a) points out, that manipulation often seems to even rely on the rational capacities of an agent (51). This is a reasonable view, even if manipulation at least introduces a non-rational influence. But does this (b) necessarily undermine our autonomy, in other words, our capacity to govern ourselves and our actions? This is a tricky question, for which I do not have enough space here, especially because there is much discussion about the notion of autonomy itself (Fischer 2017 and Fischer and Illies 2018 for more details). What we can say, though, is that an agent is influenced in a way that tries to arouse a pre-rational impulse, which leads the agent to find a specific end appealing, making her act on the grounds of that affective judgement – not because of good reasons. However, this does not necessarily degrade her to the level of an object incapable of making her own decisions. Even within the realm of these influences, an agent is still able to decide rationally and freely to not follow the paved way of a modulated affective state. Free agency thus remains robust.⁹ This, by the way, could be why using manipulation is especially interesting in liberal societies: we are still acting on the grounds of our own affectivity, usually being able to decide in favor or against it – even though this is not always an easy thing to do. Manipulation might make it harder to decide rationally; it might nudge us into a certain direction whilst flying beyond our ‘rational radar’, but it does not establish a one-way decision-making process. Every character in the examples above could have acted otherwise. Manipulation to different extents colors their rational deliberation but hardly subverts it completely. If it did, we should speak of coercion not manipulation.

4. A new integrative suggestion for defining manipulation

4.1. What makes us act?

Thus we cannot simply define manipulation as the vitiation of human rationality, our freedom and autonomy, and also if we cannot simply define manipulation as deception or the harmful means to achieve negative and selfish-ends as a manipulator. Manipulation prods an agent to choose something (relatively) freely, but the agent does not do this primarily on the grounds of reasons. How then? Generally speaking, we could say: the introduction of a pre-rational influence on the decision-making process directly modulates the target’s affective states which naturally come with the evaluation of certain ends. By curating the affective significance of an end, an affective response that is linked to the manipulative attempt, might create a desire or an aversion to do or not do something. The desire/aversion itself aims at the alteration of the reality so that reality accords. This desire might come in many forms (like anticipation, relief, drives, lust, urges, impulses, longings, cravings, … ¹⁰) as well as the aversion (like avoidance, aggression, anxiety, sadness, …) and is a foundation of a motivation. Ben for example starts to fancy taking over from Tom on the grounds of curiosity, the fear of missing out, and the promising anticipation of pleasure. In the long run, there might even be a change of belief because our beliefs are often founded in our affectivity.

Curating an end as pleasurable is one of three choice worthy ends that generally make us act. An end, according to a modernized account of Aristotle and Aquinas, (1) is chosen for its own sake; these are ‘ultimate’ and ‘fitting’ ends that regard an agent’s life, e.g. the end to become a vegan because nature and especially animals are ends in themselves (these ends have been the focus of most ethical discussions around autonomy); (2) another end is chosen because it is useful to achieve other ambitions we have, e.g. to eat vegan because it is thought to be healthy; (3) or, an end is chosen because it provides us with welfare and pleasure, e.g. being satisfied in different regards (Vogler 2002, 32; Fischer 2017, 70; Fischer and Illies 2018, 35).¹¹ If we ask ourselves how we can influence someone in a certain way, we can target the ends this person has, we add a new choice-worthy end to already existing ends, or we create a context in which one of his choice-worthy ends becomes more easily realized. Thus we can present an end (e.g. buying something or taking over the whitewashing) as (a) being choice-worthy for its own sake, which is mostly done by giving reasons as to why something is good – we rationally persuade someone with good reasons; we can also present an end (b) as useful because it helps to achieve other ends you already have. This happens often in economic contexts where a seller offers a product at a certain price and we have to decide whether or not to buy it at this price (e.g. the better sound-system or the leather seats in a new car for longer durability). And last but not least, – and this is where it gets interesting for understanding the how of manipulation – we can present an end (c) as welfare-enhancing and pleasurable and introduce a new pleasure or make some existing end more appealing and thus more likely to be chosen. All of these ends can also be presented vice versa as unfitting, useless and unpleasant.

Consequently, there is the possibility of direct or indirect manipulation and sporadic as well as systematic manipulation. Direct manipulation can, e.g., be found in interpersonal relationships where we might be manipulated directly via the induction of guilt (nana), Iago’s use of jealousy, Tom’s pitching whitewashing the fence, George’s knowing Cynthia’s soft spots (here it helps the manipulator to know the manipulated). Indirect manipulation works e.g. via advertisements that make their way into our unconscious over time with the presentation of a product in an appealing light and their constant recurrences or specific choice architectures which use i.e. our slothfulness like online 1-Click payment tools. What connects them, is the attempt to modulate our affective states in order to change an evaluation of an end (an action, a product) and create the grounds for a desire/aversion to form and make us act in a certain way (e.g., going for something or dismissing it) and hereinafter alter reality.

4.2. A new definition of manipulation

What I suggest then as a description of the mechanism of manipulation is that a manipulative act intentionally modulates the affective states of an agent by presenting a certain end (an action, a product etc.) as pleasurable/unpleasurable. This entails the suggestion to choose or not to choose this end, but where the manipulated remains, at least in a minimal sense, free to act in this manner or not. The active (though often not fully conscious) curation of the affective attraction of certain ends or their realization makes some options more (or even extremely, whereas others not at all) appealing to the manipulated and, consequently, more likely to be chosen or not.¹² As our affective states (feelings, emotions, moods) are actively modulated, the evaluation of a certain end changes, resulting in often (but not always) complex affective experiences which eventually boil down to a desire or an aversion, a yes or no regarding a certain end. This motivates the manipulated to act in a manner according to the manipulator’s goal.

By defining manipulation like this, we reach a neutral and integrative account that leaves room for the different discussed aspects and thus also the everyday language use of manipulation (Wood also pleads for a neutral account, 2014, 19). This account can but does not have to include sly and underhanded and/or deceptive manners, harm, selfish manipulators, and a damage to rationality. With this more differentiated definition, we located manipulation on the map of influences by distinguishing it from other forms of influence such as rational persuasion and coercion. This definition also avoids giving in to a strict and idealized dichotomy of rationality and affectivity and breaks the amalgamation of manipulation and devilishness.

This account of manipulation extends concepts of precursors like Baron’s (2014) notion that manipulation plays upon emotions, uses pressure to acquiescence (which is not yet coercion) or weaknesses of character (127), complicating Noggle’s and Barnhill’s examples of using guilt as well as Marcuse’s observation that manipulation works via systematically inducing libidinal needs. All of these examples are indeed central methods of manipulation which the definition above elucidates. This account can explain why manipulation does not deprive an agent of his freedom to choose: making an end pleasant might render a decision against it unpleasant and demanding, but no one is forced into a certain decision. Just consider advertisements which are aimed at a broad audience. Some agents might be weaker and thus more prone to act according to the promise of welfare and pleasure, whereas others are not at all prone and wave aside the manipulative impact of an advertisement right away. Further, someone other than Ben might have reacted differently to Tom’s attempt to manipulate. The same is true for the additional examples. Manipulation will affect – but does not utterly make – a decision.

Furthermore, this model makes it clear why manipulation cannot simply be equated with deception, because there are many different ways in which one can change the affective attraction of ends without being deceptive (even if deception might be present, or even helpful toward the manipulator’s end, in some cases). It is also important to mention that deception is not exclusive to manipulation: rational arguments can be deceitful, too.

This paper made another step in explaining what happens while we manipulate each other in an action- and emotion-theoretical manner. In the end, it is not just a matter of coming to a clearer understanding of a term and phenomenon we often bandy about – and incorrectly as I would suggest. By clinging to a negative, even caricatured view of manipulation (as essentially deceitful, imperceptible, coercive, harmful, and carried out by devilish masterminds), we are perhaps allowing ourselves to overlook the way in which we participate in the broader phenomenon, the we in which we are ‘complicit’, as it were. But complicit in what? What causes us to overlook our own use of affective states to influence others? Are we possibly too invested in a rigid dichotomy between rationality and affectivity, perhaps unwilling to consider the complex and dynamic ways they interact and act upon each other – as that could be confusing (and hard to conceptualize)? Is our hesitancy, in other words, to accept a broader, more nuanced view of manipulation an indication of an investment in outmoded, exaggerated notions of rationality and autonomy? And/or is it a flight from acknowledging the huge role that manipulation plays in contemporary politics, our relationships and commerce? There are still a lot of questions left regarding this fascinating phenomenon not just in terms of its exact workings but also the role it plays in our lives.

Acknowledgements

The author owes gratitude to William Collins Donahue for his very helpful comments, Vittorio Hösle and Mark Roche (all from the University of Notre Dame) for their insightful remarks following a lecture during my time as a visiting professor at the University of Notre Dame, Christian Illies (University of Bamberg) for the inspiring conversations and the (ongoing) collaborations on the topic of manipulation, Eleonora Meister for her perusal and annotations, Lily Carr for her advice regarding the language of a first draft, as well as the anonymous reviewers and the editors for their valuable, respectful and constructive suggestions. I also would like to thank all those whose names I don’t know but who commented on earlier versions and parts of this manuscript which I presented at various conferences.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Alexander Fischer

Alexander Fischer is an Assistant Professor of Practical Philosophy at the University of Basel (Switzerland). He received his doctorate from the University of Bamberg (Germany) and has been a Visiting Scholar at Duke University (US), the University of Cambridge (UK) as well as a Visiting Assistant Professor at Bond University (Australia) and a Visiting Professor at the University of Notre Dame (US). In addition to his academic work, Alexander is a trained psychotherapist.

Notes

1 In my understanding, we can divide our affectivity as a whole into these three categories which are all related to one another but not quite the same. Feelings are qualitative bodily impulses that can be very basic just like pain. Pain is clearly a feeling but not yet an emotion. Emotions are more complex because they contain not only feelings, but also other components of a cognitive, evaluative and motivational nature. They are intentionally related to an object in the environment and usually acute (like anger) or extensive and persistent (like a very complex emotion like love). Love, on the other hand, despite its persistence is not a mood like melancholia (which can also be constant), because it is more specifically related to an object, i.e. the loved one, while a mood has a generalized scope and colors our lives in many areas (Ben Ze’ev 2001). Consequently, feelings, emotions and moods can become anchor points in specific attempts to manipulate.

2 After various rather scattered contributions and an edited volume on the topic (by Christian Coons and Michael Weber 2014), I recently presented a definition of manipulation in the context of an elaborate theoretical setting: Manipulation. Zur Theorie und Ethik einer Form der Beeinflussung (2017).

3 Mark Roche made an interesting observation concerning Othello that I want to share here. He claims that the fact that Othello is so manipulated can partially explain why the work is less successful as a tragedy than some other Shakespeare works: Othello's suffering derives at least partly from weakness, not greatness, as in a true tragedy. At the same time, it is interesting how often in comedy, where subjects control events far less than they think, positive manipulation plays a significant role (just think of Shakespeare’s Much Ado about Nothing).

4 Rationality is usually understood as the ability of agents to be able to recognize, develop and correct reasons that will lead to a belief, action or attitude. Rationality is fundamentally committed to the principles of logic and therefore different from believing e.g. in religions. It is usually also understood as being on the opposite end of a justification-spectrum to emotion (e.g. Gigerenzer 2014, 1380). But here, it already gets messy: emotions are often supposed to be rational, too. Just think of fear while facing a dangerous animal: fear is supposed to make you run for your life and is therefore quite rational (following the logic of surviving), even though it is not the youngest and most intellectual part of our brain, the neocortex, that makes us run but rather workings in our (from an evolutionary perspective) rather ‘old’ brainstem.

5 Even though our affectivity seems to have a bad reputation, we can find accounts that attest our affectivity a share of an integrated rationality. Aristotle famously thought that our affects are rational when they show themselves at the right time, for the right amount and the right reason, e.g. grief when a loved one died. Robert C. Solomon, as an example for a modern-day emotion theorist, also conceptualizes our affectivity as a vital part of rationality as it provides meaningful judgements, discerns value, is trainable, functions as an engine of our actions, is strategic, and creates meaning (2001). In consequence, this also means that our affectivity does not in every case and necessarily undermine our autonomy.

6 Deception seems to be the paradigmatic condition when it comes to talk about manipulation, especially in the context of advertising – on the scientific side of marketing but also in public discourse.

7 Interpersonal relationships are a realm where the asymmetry caused by different resources like body strength, knowledge of the other (or a specific issue), emotional stability etc. often becomes real. Someone skilled at manipulating (it is said that e.g. narcissistic personalities are especially skilled manipulators) plays, e.g. on a weaknesses of another person and thus gains the upper hand. I recently published a paper on manipulation in the context of the Catholic Church sexual abuse cases that looks into this a little more (Fischer 2021). It could be said, that the same goes for a system like capitalism which is embroidered in our lives and where it is sometimes hard to resist against its temptations (even though not impossible – but sure: you need the resources for that).

8 The accusation consists of undermining our rationality and thus autonomy, which results in an undignified treatment of agents; in Kantian terms: only using them as a means and not treating them like ends in themselves.

9 There is the danger of overestimating the effects manipulation might have. The effectiveness can in fact vary strongly in different cases of manipulation. Whereas a hermetic relationship or a hermetically sealed state offers different scales of effectiveness, we know from other examples in our everyday lives that it is sometimes quite easy to resist to a manipulation attempt. It seems like the strength of a certain manipulation has to be evaluated case by case.

10 All these shades and forms of desires can be differentiated: A drive is a desire that arises from the body like a sex drive, an urge is a drive which is very pressing, a craving is a longing that needs to be satisfied otherwise it becomes unpleasant and so on.

11 To be sure: This is an ideal model. In our real-life these different ends might blend into each and for the simple reason that we are dealing with humans here who have a way of withdrawing themselves from the accuracy that accurate sciences would love to find in them.

12 Christian Illies and I call this concept the ‘Pleasurable-Ends-Model of Manipulation' (see Fischer and Illies 2018; Fischer 2017).