[An overview of Kegan’s subject-object theory in terms of differing expectations. Worldview differences play a key role in disagreements. Using Kegan’s stages to model social relationships and expanding compassion.]

As far as fads go in the rationalist community, Robert Kegan’s constructive developmental theory and stages of human psychological development is so 2017. Right now the cool thing is triangulation, or circling, or tulpas. I’m not quite sure. Anyway, I finally got a chance to read Kegan’s In Over Our Heads, and I quite enjoyed it.

Moreover, I think that what I got from the book was quite different than what others have written about. When I see mention of Kegan, I feel like it often gets cherry-picked as a framework which enables further Bad Things. There seems to be this pervasive meme that people use Kegan’s stages of development as a way to feel superior to others.

As a strawmanned example, someone might say, “Oh, well you’re only on Kegan’s third stage, so of course it’d seem like that way to youBut me, I’m on the fourth stage, and I see it like this…”

(I’ll explain the stages briefly later on.)

I think that everyone engaging with Kegan’s ideas has actually been way too focused on specific levels in the stages of development. And the stages, as well as the transitions between them, weren’t actually the main takeaways for me. Okay, yes, subject-object theory is actually pretty cool, and it has several properties which I think rationalists will find shiny. But I think that there’s a few interesting lessons that you miss when you only focus on how to level up from one stage to the next.

Here, I’ll try to present a different reading of Kegan’s work:

First and foremost, I see Kegan as writing about the conflict between people and the roles they play. One way to characterize how this sort of conflict arises is by explaining conflict in terms of differing expectations of our internal states.

Adapting from Kegan’s opening example, let’s think of a teenager, we’ll call her Alice. Let’s say that Alice arrives home late one night after a party:

Alice’s parents will be worried and say lots of things reprimanding her (EX: “How could you do this? Didn’t you know that we were worried sick?”).

Alice will in response try to play down her behavior (“I’m only 15 minutes late! We ran into traffic on the way back!”)

And thus the argument goes.

Kegan argues that there’s a fundamental divide here which prevents both sides from fully communicating.

Alice is thinking in terms of material gains and losses. Her relationship to her parents is an extrinsic one which has its set of rules and consequences. The sample responses she might give in the above example are ones which try to answer the question of “How do I get into the least trouble for coming home late?”

In contrast, Alice’s parents seem to want something different. They want Alice to intrinsically value the relationship she holds to them, in a manner different than just gains and losses. In their ideal world, Alice would instead be thinking something like, “My parents would be very worried if I came home late. How could I do that to them?”

This difference in what both sides expect from their relationship forms the crux of what Kegan sees to be a central point in causing conflict across the many social roles we play, from parent to child to employee to partner to leader. In particular, when the social expectation for a role is that you value X, but you actually value Y, problems arise.

For me, this framing of conflict is friendly to existing thoughts I have about things like the illusion of transparency and how communication often ends up being a weak link. Also, it shifts the question of “Who is correct in this argument?” (if we refer back to the question of Alice and her parents) into “What does each side expect from the other party and their role?” which gels well with the concept of reductionism.

To clarify a little: Kegan paints these differing expectations as conflicts of an ontological nature. When the two parties interpret the same event in different ways, it’s a question of valuing different things, of having different ideas of how roles and selves are supposed to interact. They have distinct ideas of what “good” or “acceptable” mean.

But we avoid directly assigning blame to either party. Kegan likes doing this; across the book, he takes care to stress that differences in these expectations and relationships don’t mean imply that one side is better or more correct.

Now, onto the actual overview of Kegan’s subject-object theory:

It goes a little like this: At each stage of psychological development, Kegan says there is some thing that your self identifies with, and there is also another thing that your self interacts with. As you move up, level by level, the old concept which you identified with now becomes a new object you can hold in your ontology. And as thing which you once were becomes something you see, the conception of your self similarly expands to something bigger.

That is, as you progress, you identify with a new thing, and the old thing you used to identify with now becomes that which you can interact with.

Let’s make this all a little more concrete:

At stage 1, you identify with your wants. Your in-the-moment desires are you.

  • EX: Babies doing baby things.

At stage 2, you see that you can interact with your wants. You’re no longer your wants, but you’ve now got a self that has the wants. Here, you might be more defined by your experiences.

  • EX: A grade school student saying, “I hate broccoli.” (Note how this is a statement about their preferences, more than just how they’re feeling right now).

At stage 3, you’re now relating to other people with their experiences. You see yourself as the haver-of-experiences. This allows you to connect with others better, and your self is defined in terms of the roles you play in connecting with others.

  • EX: A son who gets her mother the newest mystery novel because “she loves detective stories”.

At stage 4, you’re seeing yourself as a person who has roles, not just as your roles. You’re able to use systems to weigh and make tradeoffs between roles. Your self is more defined by the system you use.

  • EX: A father who turns down his daughter’s request to play on the Xbox because he’s filing taxes, which is “more important”.

At stage 5, you’re a collection of systems; maybe some systems are better to use at different things, but you’re not beholden to any one system. This allows you to have multiple internally consistent systems which are seemingly contradictory to one another.

  • (I’m not fully sure I understand stage 5, so I’ll let you fill in the blank here for an appropriate example.)

As a result of reading about these different systems of valuation, I’ve noticed myself being more curious in conversations, asking myself “What am I vs the other person implicitly valuing in this moment?”

To that end, I’ve also been thinking about the nature of relationships of all kinds lately, and Kegan does a good job of painting how the reasons you have to care for someone else in a relationship, changes as you go through the stages.

A question that I’ve been asking myself is “What are the factors that lead me towards liking someone / becoming friends with them?”

Here are some reasons that might lead you towards forming a relationship with someone:

  1. They have something you want.
  2. You have shared interests.
  3. You care about them as a human, even if you have different interests.

For me, I’ve been consistently using shared interests as a major filter for finding people I’d want to spend time with. Sometimes, though, I also end up liking people who are quite different from myself. But if we aren’t bonding over shared relationships, than what’s going on here?

Kegan’s explicit model of how these different motivations for relationships form as a result of differing stages has helped me get some more clarity in this regard:

  1. Thinking about those around you in terms of direct gains/losses is representative of the consequential viewpoint in stage 2.
  2. Wanting to share in mutual experiences and interests with those around is representative of the experiential viewpoint in stage 3.
  3. Valuing other people as independent entities with their own goals and interests is representative of the autonomous viewpoint in stage 4.

Looking at the different relationships I hold with others, though, I notice myself holding a mix of all three of the above reasons (though I do seem to bias towards shared interests, as I mentioned above).

I found this valuable because it reminded me to care about others for reasons more than “we get along well”. I think exercises in expanding value are good, and the expanded field of acceptance fits in well with what I wrote earlier about not being judgemental.

I’d like to think of the sorts of relationships you form at different stages as all being viable reasons to connect with someone. The dependent factor isn’t what level of psychological maturity you’re at; it’s context-dependent:

  1. Networking can be thought of as a stage 2 relationship that pops up all the time, with both sides offering to exchange resources.
  2. Dating often takes the form of a stage 3 relationship, with both parties looking for overlaps in values to bond.
  3. Good friends eventually become people you care about, independent of the initial reasons that brought you together, in a way that’s much more like a stage 4 relationship.

Using this framework as an investigative tool has revealed how I’m often transitioning between different sets of values with different people, and sometimes even with the same person. So it looks like there’s several forms that our relationships can take. It’s also precisely for this reason that I don’t like viewing Kegan’s theory as one of direct progression.

Even as you mature, there are going to be situations where the assumptions/expectations of a previous stages form a valid basis to hold a relationship, depending on the circumstances. This idea of having multiple frames, of being able to switch between different viewpoints forms the basis of the idea behind fake frameworks I think it’s also largely representative of the “system of systems” approach that Kegan outlines in stage 5.

If you start using Kegan’s stage 5 to look at his entire theory, it feels less like a progression of development (although there is indeed a view where you can cast it as such) and more just like a set of different lenses for your mental models toolkit.

Overall, I’ve found In Over Our Heads useful for additional intuition pumps to feel compassionate towards others, as well as a model for thinking about the different bases for relationships.

New Comment
4 comments, sorted by Click to highlight new comments since: Today at 5:47 PM

For me your examples fall a bit flat because they don't sufficiently address the internal experience of the subjects and how they are relating to the world. Just to deconstruct one of them:

A father who turns down his daughter’s request to play on the Xbox because he’s filing taxes, which is “more important”.

Well this behavior could be evidence of different stages on the father's part depend on what it means to him that filing taxes are more important:

  • If he thinks taxes are simply a more important thing than playing Xbox that's more indicative of 2.
  • If he thinks he has a responsibility (implicitly to society, his family, or something else) to do taxes and the source of that responsibility values doing taxes over playing Xbox then that's more indicative of 3.
  • If he is aware of his daughter's desire to play Xbox but weights this against his own desire to do the taxes and decides doing the taxes is more what he wants, that's more indicative of 4.
  • If he considers his daughter's desire to play Xbox, his own desire to do the taxes, considers the consequences of choosing one over the other for himself and for his daughter (and maybe for his family and society more generally), and decides that the consequences for his daughter, himself, and others of his doing the taxes rather than playing Xbox suggest that he can most get the world to look like he wants it to if he does the taxes and the best way to make this legible to other is to say that the taxes are "more important", that's more indicative of 5.

(extra details in the 5 example since it's the area you expressed most confusion over)

The point is that developmental psychology often addresses behaviors that cannot be easily observed externally because the behaviors being changed are patterns of thought rather than actions in individual cases. To put it another way, if faced with a trolly problem being at a higher developmental stage won't necessarily change your choice, but it will change how you make the choice.

[-][anonymous]6y30

Ah yes, you're correct about how the internal experiences differ. I don't think I did a good job of pointing out how that is the thing that differs.

Thanks for the additional exposition on stage 5!

Thank you for this description; it felt useful! I think you are approaching this the right way.

It reminds me a bit of how I use Freud; some ideas seem roughly correct, so I take them as a framework, but I don't get obsessed over details. For example, I accept that there are multiple agents in mind, but I don't think there are exactly three of them, nor that they exactly correspond to the "id, ego, superego". But the idea of internal conflicts and internal forces outside of my attention or control, seems supported by experience. I specifically like the descriptions of defense mechanisms; not because I would think the list is complete, but because having names for multiple frequent patterns feels more useful than merely having a vague category called e.g. "irrational behavior".

In other words, it's not a package deal. You can learn from a person without copying their whole map.

Similarly, the lesson I take from your article is that there are multiple patterns of how people are seeing themselves, and what they expect from others. And those patterns seem to be ordered from simpler to more complex. Doesn't mean the list is necessarily complete. Doesn't mean that each specific experience needs to fit into exactly one category. Doesn't mean that people must use this categories linearly, exactly one at a time. Okay, I would in general expect that the more complex categories are learned later in life, but I suppose that people who learn the new ways are still using the old ones quite frequently. Or that they compartmentalize in various ways, e.g. follow one pattern more frequently at home, and another one more frequently at work.

Maybe some experiences, such as playing role-playing games, allow a person to progress atypically, e.g. they may gain the insight of "playing a role" much faster than would correspond to their overall level of maturity. Actually, I would expect people to vary in their "Kegan levels" depending on whether at given moment they feel hungry, angry, lonely, or tired. That people are more likely to identify with their emotions at the moments when the emotions are strong. Or more likely to recognize a role as a role in situations where the role has a clear label and rituals for turning it on and off, such as wearing a uniform; so e.g. a policeman would be aware that "policeman" is just a role they sometimes play, but might not realize that e.g. "parent" is also a role.

As a strawmanned example, someone might say, “Oh, well you’re only on Kegan’s third stage, so of course it’d seem like that way to you…But me, I’m on the fourth stage, and I see it like this…”

Such person would obviously be on Kegan's third stage, because they identify with a role. Even if it is a role of being, in their imagination, on a different Kegan stage. /s

I'm confused by your description of how you orient towards relationships with others (among other things, it doesn't come close to carving my experience at the joints), but I'm not sure how to articulate the confusion.