Note: this is from my personal blog here.

My sister (hi Lindsay) and I are very different people. Give Lindsay an hour with 15 people, and she will introduce you to her 15 new best friends. Give me an hour with 15 people, and I might remember one of their names. Need a mediator for some sort of drama? She’ll have everyone hugging (even over Zoom) after a heartfelt and compassionate discussion, whereas I’m more likely to take sides with whoever seems to be in the right.

In psychological terms, these differences boil down to two of the Big Five personality traits: extraversion and agreeableness. On a scale from 1-10, Lindsay is probably a 10 on extraversion and an 8 on agreeableness. I’ll plead the fifth instead of rating myself on both (and I don’t think I’m a mean person, honest!), but I probably do fall a bit lower than both of those.

What is agreeableness?

As shown in the picture above, agreeableness has six “facets” or features, each of which has a section in the Wikipedia page:

Trust
Straight-forwardness
Altruism
Compliance
Modesty
Tender-mindedness, or “the extent to which an individual’s judgments and attitudes are determined by emotion”

A missing definition

Surprisingly, I can’t seem to find a unique definition of agreeableness as used in psychology. Here are some quotes I came across:

The picture above defines it as “the degree to which a person needs pleasant and harmonious relations with others.”
“A person’s ability to put other people’s needs above their own. For instance, people who are high in agreeableness naturally experience empathy and tend to get tremendous pleasure from serving others and taking care of them.” (Source)
“Agreeableness is a personality trait manifesting itself in individual behavioral characteristics that are perceived as kind, sympathetic, cooperative, warm, and considerate.”(Source)
“Specifically, Agreeableness appears to describe differences in being predominantly prosocial or other-oriented versus antisocial or self-oriented in social interactions.” (Source)
“Agreeableness reflects the individual’s tendency to develop and maintain prosocial relationships.” (Source)

There is always a risk of being pedantic or counterproductively hyper-analytical when trying to define a term. Don’t these five definitions and the six facets cluster around the same intuitive phenomenon?

In part, yes. I can easily imagine examples of unambiguously agreeable or disagreeable behavior. But it gets interesting when some of these facets or definitions come into conflict.

Let’s take a look at those definitions again.

Definition 1 regards agreeableness as a personal preference or desire, just as some people might have a relatively strong “need” for relaxation, intellectual stimulation, or exercise.
Definition 2 regards agreeableness as a skill or strength, like intelligence or physical fitness.
Definition 3 defines agreeableness as how a person is perceived, rather than as an intrinsic characteristic. Thus, the same behavior might be agreeable in one society and disagreeable in another even if the person and intent are the same.
Definition 4 regards agreeableness as a cognitive disposition that makes some things—namely, other people—more salient or important.
Definition 5 is behaviorist, defining agreeableness as a tendency to do certain things in the world.

But before returning to the facets and definitions: a detour…

Into other frameworks

We can fit agreeableness into several existing analytical frameworks. From a game-theoretic perspective, agreeableness is something like a tendency to cooperate instead of defect or exploit others. Sometimes, this is great! Play an iterated prisoner’s dilemma, and everyone is collectively best off when all agree to cooperate. Unfortunately, people don’t always cooperate, so an agent needs a credible threat of defection in order to deter its partner from doing the same.

A bit further from the hyper-simplified world of mathematical abstractions is Scott Alexander’s Conflict vs. Mistake Theory dichotomy. Some people tend to view problems as conflicts between individuals or groups with fundamentally opposing interests, and other folks tend to view them as unfortunate, de-moralized circumstances—mistakes—which can be improved for the benefit of all.

Obviously, neither tendency is “correct,” and either model can be a better fit for one circumstance or another different angles of the same situation. Trying to understand basketball as a “mistake” in which all the players could agree to work together to win isn’t very helpful, but neither is viewing the economy as a ruthless, zero-sum competition for resources.

To play Enlightened Centrist for a minute, though, basketball is partially a positive-sum situation, and economic activity is partially zero-sum. Both teams can work together, in a sense, to simultaneously develop their skills, have fun, and entertain the fans. And, at any given time, each unit of economic resource can only be controlled by a single entity.

Agreeableness aligns closely with mistake theory, and agreeable individuals act as though problems have positive-sum solutions. That’s why Lindsay would be such a good mediator. Unfortunately for agreeable people, conflict theory is sometimes the best model of a situation; in a fundamentally zero-sum competition, you can’t coherently “agree” with everyone—you have to pick a side!

When facets face off

Trust, straight-forwardness, altruism, compliance, modesty, and tender-mindedness do not always fit well together. Here are some ways they can come into conflict:

Altruism vs. Compliance

This pair, stands out to me as the one with the most conspicuous potential conflict. When someone is being instructed to harm others, there is a zero-sum tradeoff between these two facets—a situation best modeled by conflict theory. There are countless examples:

A soldier is instructed to kill civilians to beat the population into submission.
An accountant is instructed to cook MegaCorp’s books to help it evade taxation.
A participant in the Milgram experiment is ordered to continue administering shocks to a fellow subject.

In these circumstances, choosing altruism over compliance seems to be the ethical thing to do. However, I can imagine other circumstances which evoke less of a clear intuition:

A judge is “instructed” by the law to sentence a pot dealer to 20 years in prison, even though the sentence seems excessive.
A participant in a real scientific experiment is instructed to administer moderately painful but safe electric shocks to a subject who consented to take part in the study.

Straightforwardness vs. Tender-mindedness

The former is “the quality of directness and honesty in communicating with others” and the latter “the extent to which an individual’s judgments and attitudes are determined by emotion…primarily defined by sympathy.”

Clearly, there are times when being straightforward and honest requires ignoring or muting one’s own or another’s emotion:

Telling your wife that, yes, that dress makes her look fat.
Giving one’s honest assessment of a poor-quality piece of writing when asked for feedback.
Telling a friend that they shouldn’t waste time applying for that job they’re not qualified for.

Trust vs. Altruism

Doing what you think is in a person’s best interest might require not taking their words at face value:

Giving a friend a birthday present after she tells you she doesn’t want one.
Reviving a seemingly young and healthy person with a “Do Not Resuscitate” tattoo (not hypothetical).

Modesty vs. Straight-forwardness

Should you honestly report an opinion that would come across as arrogant?

Replying “I’m smarter and harder working than most other people” when asked about the keys to one’s success.
The best player on a soccer team saying “pass the ball to me” when asked what the team should do to win.

With a little more creativity, I’m sure I could come up with examples for all 30 combinations of the six facets, but you get the point: (dis)agreeableness doesn’t always prescribe a course of action.

Gestalt

Despite the heterogeneity of agreeableness’s definitions and descriptions, I—and I suspect most of us—have a sort of intuitive, unified impression of the trait. Since the Big Five seems like a pretty robust scientific model, I’ll operate under the assumption that agreeableness reflects a real, underlying psychological phenomenon that unites those six facets.If this is correct, here is my attempt to unite those five different definitions.

Definition 4 gets it closest: agreeableness is something like “other-orientedness,” or the degree to which whatever things in one’s ontology are understood as fellow people are salient and important. This would manifests in one’s:

Personal preferences, the desire to please others. This arises because others’ preferences/emotions constitute a larger proportion of one’s own.
Skills, the ability to please others. This arises because of greater practice and desire to do well, and is likely a consequence of personal preferences.
Impressions on others. Those who care more about pleasing others and are better at doing so will of course tend to make better impressions.
Behavior, the outward manifestation of one’s preferences. This is also reinforced by positive feedback due to superior social skills and impressions.

Under this understanding, it is a little easier to guess how an agreeable person will behave when facets conflict. Often, his or her behavior will depend on the visibility and “personhood-ness” of the people/groups/entities/beings with competing interests.

For example, the accountant instructed to evade taxation will see a real, live human instructing him to do so, and is likely friends with others who are participating in the scheme. The victim, on the other hand, is the abstract notion of one’s fellow citizens, the IRS, or “rule of law.” Since most of us regard a boss or colleague as a fellow person but not one’s countrymen at large or the IRS, an agreeable person will be more likely to participate in the scheme.

On the other hand, the situation is flipped for the judge “instructed” to sentence a pot dealer to 20 years. She can see the defendant in front of her in the courtroom, and surely regards him as a person. The letter of the law, on the other hand, is a conceptual abstraction. Therefore, if they indeed have some latitude in reducing this sentence, an agreeable judge will be more likely to be noncompliant.

I selected these two examples to be morally offsetting. An agreeable person is more likely to do the (intuitively) immoral thing in the first situation and the (intuitively) moral thing in the second. When I think of agreeable people whom I know, this just seems to “fit” with what I’d imagine them (relative to base rates!) likely to do.

I imagine my sister (don’t worry, she approved this post) more likely than me to go along with the tax evasion scheme (in a distant universe in which she was an accountant), but also more likely than me to give the pot dealer a lenient sentence.

Is agreeableness good?

Compliance is wonderful when authorities are competent and ethical. Trust is fantastic when people are honest. A willingness to go with the flow is great when the flow is going in the right direction.

But disagreeableness does have its virtues. Or, as a disagreeable person might put it, agreeableness has its problems. Authorities are sometimes incompetent and amoral, people sometimes lie, and the flow is sometimes going the wrong way. The Holocaust is the most tragic such example, and the Milgram experiment (which does replicate!) provides a microcosmic demonstration of potential consequences of trust and compliance—in a word, agreeableness.

Cooperation or Complicity?

When our or our community’s actions produce negative consequences for others—whether intentionally or otherwise—agreeableness looks more like complicity. Returning to the previous two examples, it seems that agreeableness would have predisposed a person to “go with the flow” of the Nazi regime or comply with Milgram’s scientist’s orders to continue shocking a fellow study participant.

No, I’m not saying agreeable people are literal Nazis. In fact, Hitler seems to have been highly disagreeable. However, a normal German citizens in the 1930s who wished to resist the Nazi regime would have had to take actions that felt belligerent and disagreeable from the inside, even though in an objective moral sense resisting genocide sure seems like a prosocial thing to do.

Both.

At the personal level, agreeableness and disagreeableness are self-limited by feedback cycles; an agreeable person who gets taken advantage of too many times will learn to adjust his behavior, and a disagreeable person who finds herself unable to make friends will likely try to behave in a more welcoming manner.

At the social level, though, this is not the case. We often make decisions or take actions that impact others. When consequences are limited to other humans in our community with a similar amount of power, agreeableness is generally good for others, even if it comes at the cost of our own narrow or selfish preferences.

If I’m correct about all this, agreeableness biases a person to behave ethically when the beneficiaries of a potential action appear to be real fellow persons. For instance, I imagine an agreeable person more likely to donate money to a developing world charity when presented with the profile of one of the charity’s beneficiaries.

On the other hand, disagreeable people are more likely to act ethically in defense of beings hidden by distance or time. This would predispose a person to give (relative to base rates!) more weight to farmed animals or people who may exist in 2893, both groups I’ve referenced in other posts, over her friends or the presently-alive poor.

I think this all becomes more intuitive when disagreeableness is viewed as a low “degree to which a person needs pleasant and harmonious relations with others” instead of an affirmative desire to be an asshole. At least, this is how I understand myself.

Humans have all sorts of goals and desires, of which interpersonal harmony is but one. Ethical aims sometimes but do not always align with interpersonal harmony, and prioritizing one goal must sometimes come at the cost of another.

Agreeing to Disagree: towards a better understanding of agreeableness