Influence and manipulation are both attempts to alter the actions or propensity to act of another agent. The only difference between the two that can generally relied upon is that the one who calls it manipulation finds it distasteful or immoral. If you disagree I invite you to find a general principle cleanly dividing the following examples into manipulation and influence; all the better if it can be made uncontroversial.

 

A woman telling her long-term boyfriend that they're not getting married before he gets his doctorate.

A man teasing a friend.

A man teasing a female friend.

A man teasing a female friend, flirting.

A man teasing a female friend, flirting with intent.

A man teasing a woman he met ten minutes ago, flirting with intent.

A woman encouraging her son to become a teacher because the job security is good.

A woman encouraging her son to become a lawyer because he'll be better able to support her in her old age.

 

I'm not denying that manipulation and influence can be usefully distinguished. I do not believe they can but I haven't spent the last month thinking about it on and off. There might exist a Schelling point dividing bidirectional communication into socially acceptable influence and unacceptable manipulation, a lawyerly thing, useful but without any defensible reason to be there exactly but that it must be somewhere and we have come to an agreement that here will do. If you believe that there is a point that actually is pragmatically better than another by a real margin, please state it, and defend your proposition.

New to LessWrong?

New Comment
48 comments, sorted by Click to highlight new comments since: Today at 7:44 PM

If I can tell you what I'm doing and it still works, it's influence.

If my telling you would make it ineffective, it's manipulation.

I think this is it exactly. But let's be rigorous:

  1. "I am trying to get you to get a doctorate by holding back marriage." Still works.
  2. No goal (that I can see), therefore nothing.
  3. No goal (that I can see), therefore nothing.
  4. No goal, therefore nothing (assuming I interpret 5 to mean that 4 is NOT with intent).
  5. "I am teasing you to get you to have sex with me." This probably will not work, but partly because it interrupts the flirting rather than because she knows that the flirting is going on. Depends on how severe the teasing is, really.
  6. "I am teasing you to get you to have sex with me." This probably will not work either, same reason as 5.
  7. Assuming no ulterior motives, "I am trying to get you to become a teacher by this encouragement" Still works.
  8. "I am trying to get you to become a lawyer by this encouragement" Okay, here's an error; it's clearly not that I can tell you what I'm doing that's necessary. Or at least, not alone.

So, with the corrections suggested by doing this, the distinction should be:

If the target having full knowledge of what you're doing doesn't affect whether it works, it's influence. If the target having full knowledge of what you're doing does affect whether it works, it's manipulation.

Or, to get at why one is immoral and the other isn't, if there's deception involved it's manipulation. If there isn't it's influence.

I had a somewhat similar thought. I was contemplating an unusual situation where society considers manipulation acceptable and influence unacceptable -- the relationship between attorneys and jurors in the adversarial legal system. Trying to subtly manipulate the jury through selective presentation of evidence or slanted wording of questions = doing your job well. Trying to influence the jury via bribery or intimidation = jury tampering.

I have yet to find anything wrong with this definition:

Manipulation is influence we don't like.

[-][anonymous]13y00

Yet surely there is something wrong with it. That's the amazement of language. It's substantially more complicated to analyze than to use. It's almost certain that one cannot explain the difference between two words with a handful of other words. The brain is composed of subtler probabilities.

I invite you to find a general principle cleanly dividing the following examples into manipulation and influence

Precise conceptual analysis being impossible doesn't make all words have the same or trivially related meaning (clearly). What is the purpose on this exercise?

I'd probably say none of the examples are manipulation by any definition I know that doesn't have all of influencing as a subset. By those, manipulation is the kind of thing used for brainwashing prisoners, or uFAI escaping from boxes, or Stanford basilisks, or drugging someone without their consent, or the like.

I have sometimes been described as slightly manipulative by others, and never quite understood why, it sounds like it might be related.

You're probably on the hard-to-manipulate end of the spectrum. Predictions: you don't tend to buy things from fast-talking door-to-door salesfolk, you don't often start liking people who upset you when you're around them, you wouldn't give anyone money or sex just because you'd be embarrassed to refuse. Many people are the other way around.

I can't remember having ever been in any of those situations, but introspective testing (unreliable, but better than nothing) says you're probably right, and roughly analogous situations indicate the same.

... and how the hay do people like that survive? That's just so... exploitable. I consider myself extremely easy to manipulate, but that's in an absolute compsci way and I have no idea how it places me on a curve compared to other humans... If what you're saying is true, then there is a fair (but still not large) chance I could make a working suicide rock!

how the hay do people like that survive?

Good question. A large part of the answer is probably that exploitation is considered manipulation by onlookers, and so potential exploiters (correctly) deem it ethical to refrain, or are punished.

a working suicide rock

Mass suicides in doomsday cults, and suicide pacts, suggest that it partially works on particularly vulnerable populations.

Good question. A large part of the answer is probably that exploitation is considered manipulation by onlookers, and so potential exploiters (correctly) deem it ethical to refrain, or are punished.

I suspect that it's often by staying within "safe" social circles in which people don't take advantage of them too hard.

Ok, that's scary... O_o

Not remotely as scary as a whole lot of other things that are also real and I know about. But still pretty scary.

And the thing you mentioned where they try to cooperate dosn't sound like it'd work very well, more like it'd just introduce even further vulnerabilities at least once you take it into account. But maybe it turns out more effective in practice.

Update: I've been playing around a bit with prisoner's dilemma simulations (not particularly ruthless ones). If nice retaliating strategies (like tit-for-tat) rise fast enough, then patches of unconditional cooperation can survive by leaning on them and each other. So it seems to works. My simulation wasn't particularly ruthless, though.

/me is a born sucker... probably

It seems to me that once we get away from obviously problematic situations (such as blackmail), the distinction is going to be in intent, which is never uncontroversial.

Agreed. But is there any point beyond expressing opprobrium in having two words if you're not really talking about anything except your own feelings? What I'm really asking is whether a behaviourist along the lines of B.F. Skinner would have been able to distinguish the two concepts. If not, it goes in the "may need to phrase this differently depending on audience but no actual difference in facts or anticipations" category.

It's been a long time since I logged into LW; I just saw this. Actually, I released a book this year in which I analyze manipulation fairly extensively through the lens of the pickup artist subculture. It's called Confessions of a Pickup Artist Chaser: http://clarissethorn.com/blog/2012/03/08/confessions-of-a-pickup-artist-chaser-now-available/

The fact that acts A and B inspire different reactions from the community (e.g., acceptance, approval, punishment, etc.) is a meaningful distinction in behaviorist terms.

Economy = greed, valour = fanatism, piety = bigotry, precision = nitpickery ... the examples may not be the best ones, I am not so good at generating synonyms of this kind. But there are lots of words which denote the same thing but are perceived as distinct because of the moral judgement contained in their connotations. "Influence" is neutral, "manipulation" is disapproving. An attempt to locate the boundary between these two either assumes common (or objective) moral position, or doesn't conserve the connotational part of the meaning of the respective words.

Well well... having been involved in the PUA community for quite a long time made me ponder this theme a lot. This is how I see it.

First some definition: regarding personal data, a truth is some fact that is true of us, and a lie is some fact that is false of us. Regarding behaviours, a truth is some behaviour that is sustainable, i.e. can be indefinitely repeated, either because it is a habit or because we have the intention to persevere in doing it. A lie is then defined as some unsustainable behaviour, that is something we do with effort without the explicit, conscious intention to repeat it.

Then, in my book, revealing a truth and omitting a lie is influence, and revealing a lie and omitting a truth is manipulation (that is, morally despicable). In table form

---------- Reveal --------- Omit

Truth ---- Inf. ------------- Man.

Lie ------ Man. ------------ Inf.

This table spurs a number of interesting and, in some cases, counter-intuitive moral tenets. For example, if I have a big car and I tell to a girl that, it's influence. Of course it's manipulation if I don't own a fancy car. On the line below, buying a drink for a girl when I don't normally buy drinks for people I don't know, that's manipulation (revealing a lie). OTOH, if I'm shy around women that I don't know and I'm attracted to, but not shy around women I know and I'm attracted to, then trying to be at ease around women in the first group amount to sustainably omitting a lie, which makes it influence and not manipulation.

Regarding the behaviour indicated by the OP: 1 - influence if she normally does that, manipulation in the other case; 2 - normally influence; 3 - normally influence, depends on the shyness of the flirter; 4 - same as above; 5 - usually manipulation, again it depends; 6 - same as above; 7 - depends on the choice career of the mother: did she choose job security? If so, influence, in the other case, manipulation; 8 - depends on the mother: if she cared about her son, influence, otherwise, twisted manipulation.

I think that most people intuitively draw the line at whether someone is doing something intentionally or unintentionally, then don't reflect on the distinction.

What one person does naturally could be manipulation for another person to try, if they explicitly learned how to do it and then decided to for some purpose.

What about when someone (a) makes a deliberate project of improving something and blunders their way into being better at influencing people through trial and error? (b) as above except their trial and error is heavily guided by some kind of theoretical framework, some system? (c) as in (b) except it's been so long since they were actively trying to improve that they've internalised the new behaviour, so that they could probably explain what they're doing if pressed but they're doing it subconsciously?

If the intuitive drawing of the line is either incoherent or makes no sense according to one's moral theory then one can ignore it, yes? Whether one is entirely open about what one is doing is another question, it's rare that that's a good policy.

I can think of a lot of lines I would defend that sort acts of influence into socially-acceptable and not, or moral and not, or pragmatic and not, or various other divisions. But I agree with you that none of those lines are between influence and manipulation (nor between manipulation and coercion, nor between coercion and force).

I wouldn't expect them to be: manipulation, coercion, and force are all sometimes socially acceptable, moral, and pragmatic.

[-][anonymous]13y00

Hm, I should find a good Jedi Kreia quote for this.

Then again, maybe not.

[This comment is no longer endorsed by its author]Reply

Hm, I'm trying to delete a double post. All I can see is this "retract" button, after which the post is crossed out but still there, even when I've logged out. What am I doing wrong?

[This comment is no longer endorsed by its author]Reply

A recent site update replaced the deletion function with the retract function, which does what you describe. With the available functionality, you can edit your comment down to nothing, or just leave it crossed out, as you prefer.

Thank you. I don't understand why this change was considered a good thing, but no matter.

The way I use the term, manipulation is influence that is improper either due to the methods used, or the social context in which they are being used. Our society (and I suspect any complex society) will have very complicated norms about what kinds of influence are legitimate in what circumstances. For instance, lying to induce somebody to do something they wouldn't do if you told them the truth is improper manipulation in most circumstances.

A lot of our norms are area-specific. For instance, in law, there's a very well developed set of principles about precisely what constitutes misrepresentation or undue influence -- in which case the contract can be void.

I suspect, in general, there will not be good reasons for putting the demarcation at one point over another: I imagine that social processes will drive norms towards at least a local cost-benefit maximum. And if the cost-benefit landscape is smooth, those local maxima will be places where there landscape is locally flat, implying that there's negligible harm if you move the equilibrium slightly.

One aspect that seems to be relevant is the level of harm / benefit to the people involved, and the degree to which the influenced person is willing to go along with it.

As for PUA - I might think that people are looking at signals of [fitness / how fun you might be to have sex with / whatever], and deliberately manipulating these (instead of the underlying traits) degrades the reliability of the signal, effectively "lying" about the quality under consideration.

On the other hand, if you think you're actually more fun to have sex with than you seem, then you might be more accurately representing yourself by changing how you seem.

If good mathematicians are believed to wear the yellowest shirts, and you think you are the best mathematician in your class, and at the moment your shirt is only pale yellow, then you would not be deceiving anyone by changing into a very yellow shirt. You might be wrong in your appraisal of your mathematical skill, and as a result you might signal a falsehood; but this would not be deception.

Absolutely true. Nerds do have certain advantages in bed!

As for PUA - I might think that people are looking at signals of [fitness / how fun you might be to have sex with / whatever], and deliberately manipulating these (instead of the underlying traits) degrades the reliability of the signal, effectively "lying" about the quality under consideration.

Do you also consider wearing make-up, push-up bras, high heels and similar to be “lying”?

Do you also consider wearing make-up, push-up bras, high heels and similar to be “lying”?

The user you are asking the question of made the grandparent in 2011 and hasn't posted recently.

Indeed they haven't. Weird, their username did ring a bell. Whatever -- it's not the first time my memory fails me like that. shrug

their username did ring a bell.

Nice.

Do you also consider wearing make-up, push-up bras, high heels and similar to be “lying”? [emphasis added]

One of these is not like the others ;)

If I get your point, I think I usually could tell whether someone's wearing make-up, if I looked close enough. And at a first glance certain shoes look less high than they actually are. Dunno about push-up bras -- I'd have to test myself by looking at women wearing them and women not wearing them without knowing which are which and trying to tell them apart, but where on Earth could I get a chance to perform the experiment? ;-)

but where on Earth could I get a chance to perform the experiment? ;-)

The most obvious solution is to form your own harem.

=P

Interesting topic. Various kinds of Dark Arts seem to fall in this influence v. manipulation gray area. I think the distinction is going to depend on context, and the projected benefits or drawbacks of said influence to the target and possible third parties. For instance, take the use of Powerpoint slides: They are often used in the classroom for convenience, yet many management experts decry their use in the workplace, because the whole setup connotes the speaker as having a higher status than coworkers, the audience usually falls into a "flow" state or semi-trance (useful in academia, very much not so in business), and the slides' contents are subconciously perceived as factual and objective even when we conciously try to maintain an open mind.

This problem is going to become more and more serious as research in neuroscience improves: the Lawrence Livermore Laboratories are already working on "cognotechnology" (possibly bordering on mind control according to Gerald Yonas) for the US military. Maybe we need a moratorium on neuroscience and ems research, to complement the SIAI-proposed ban on unfriendly AI development.

If you believe that there is a point that actually is pragmatically better than another by a real margin, please state it, and defend your proposition.

I'm highly unusual, but seeing as I've optimized my choice of friends pretty well, for me personally there's a pretty clear point of divergence, and that's the point of divergence that causes some commenters on Less Wrong to get all worked up about PUA. Seeing as it gets those commenters all worked up it presumably would get some non-negligible fraction of less reflective smart people all worked up as well, so perhaps it extends beyond just Less Wrong. In the majority of cases I've heard discussed, if something isn't explicitly-learned-and-consciously-acquired attracting-women-in-optimized-ways then it's probably influence, if it is then it's obviously manipulation. Yawn-inducing but practically useful in a way that a thorough understanding of the social psychology of manipulation wouldn't be. ("Manipulation" just never seems to come up. Is this not true for most folk here?)

(In case it's not blatantly obvious, yes I agree the existence of this Schelling point is unfortunate, no I am not going to waste time complaining about it or dreaming about marginally better ones.)

In the majority of cases I've heard discussed, if something isn't explicitly-learned-and-consciously-acquired attracting-women-in-optimized-ways then it's probably influence, if it is then it's obviously manipulation.

The problem with using that as a Schelling point, is that it means that influence can become manipulation simple by virtue of the influencer becoming becoming more self-reflective and thinking about what he's doing and why it works.

I think that's basically too far outside of the model of the sort of people who use that Schelling point for for the consideration to occur to them, so they don't realize it's a problem.

Self-reflection is far outside their model?

The model they are using is of probably-morally-contemptible PUA folk. Morally contemptible people are morally contemptible. Good people don't become morally contemptible. If someone becomes a PUA-esque person after engaging in reflection then that just means they were morally contemptible beforehand as well. And morally contemptible people don't engage in self-reflection anyway so they'd they'd never end up as a PUA-esque person. People who engage in self-reflection properly will look like fuzzily imagined knights in shining armor, because they are not morally contemptible, obviously.

This only applies to vaguely imagined PUA folk. If a PUA-esque person actually shows up and starts talking to a person with this kind of model then of course it no longer applies.

The above is my general impression of the more saddening parts of what I perceive to be the kind of implicit social reasoning involved. Others will have better general impressions.

So don't learn about social functioning from those who have expertise. If you do learn from personal experience don't theorise it or systematise it, if you have done that don't talk about it?

Sounds about right for the playing of social games generally, doesn't it? How depressing. Lionhearted's new post seems relevant, particularly pjeby's discussion

If you do learn from personal experience don't theorise it or systematise it, if you have done that don't talk about it?

No, that doesn't matter either, I think. Individuals who publicly identify as being influenced by PUA thinking can be seen as virtuous given some moderate level of social savvy. It's only the vague class of imaginably contemptible PUA folk that is manipulative.

Would it be fair to say the disdain is mostly signalling then? Moderate social savvy is not a high bar.

That depends a lot on the specifics of the connotations of "signalling" I think. The people who engage in such signalling always have the best of intentions and obviously are only trying to positively influence others, never manipulate them. (Am I insulting them by insinuating that they are unreflective, or being charitable by saying they're not consciously manipulative? Or am I being sarcastic? Does it matter either way? You decide!)