The Cofounders

Prithi and Carlos
Image courtesy Sajjad Hussain M

Meta-systematicity manifests as the forefront of all domains of meaning, including in personal psychology, rational understanding, social organization, and culture. Observing meta-systematicity across these domains reveals structural parallels, so that insight from each domain illuminates the others.

This page braids together three purposes:

The topic poses a chicken-and-egg problem. The ability to think meta-systematically is a prerequisite to learning what meta-systematicity is—just as the ability to think systematically is a prerequisite to learning what systematicity is.1 So the path from systematicity to meta-systematicity is difficult, gradual, and takes years.

A map of the way ahead helps. Learning about how one learns to be meta-systematic may be the best way to learn what it is to be meta-systematic.

I illustrate the path to personal and relational meta-systematicity with a series of snapshots of the development of a fictional character, Prithi. She is the CTO (Chief Technology Officer) of an imaginary technology startup. She cofounded it with her college friend, and now co-CEO, Carlos.

Each snapshot features a monologue in which Prithi explains how she understands some problem in a fictional interpersonal situation. Each story is simple and concrete, to make understanding easier. However, it would still be easy to miss the points, which are subtle and abstract. I will ask you to listen to each monologue in a peculiar, difficult way: at three levels simultaneously. There are the events: who said what. There are the meanings Prithi ascribes to those events. And there is the way those meanings come to be, and how they interact with the events, characters, and each other. The first two levels, the events and their meanings, are simplified specifics I invented to give substance for the third, which is the topic of this page.

The third level, the dynamics of meaning, progresses from systematic to meta-systematic. In the systematic mode, meanings are structured by a collection of rules, policies, principles, and procedures that interlock as a coherent system. Systems can function extremely effectively, but are brittle in the face of nebulosity—uncertainty, change, ambiguity, and indefiniteness. The meta-systematic mode addresses nebulosity by expanding concern outside any system, to consider how systems relate to the unspecifiable details of reality, and to each other.

So what matters is not so much what Prithi says, as how she comes to say it. Listen for incremental shifts in how she, systems, and meanings relate to each other. This may require repeated re-readings. I will try to help: with analysis after each monologue, and in some cases by interrupting Prithi to comment on what she is saying in real time.

For understanding relationship dynamics, the story context of business leadership is just set-dressing. How Prithi thinks and feels and acts could be illustrated equally well in the context of a marriage, for example.2 (In fact, the cofounder relationship is often likened to one.) However, the operation of a business or other organization can also shift from systematic to meta-systematic. Prithi will describe this change in the final snapshot. Then I’ll say more about meta-systematic organizations in an epilogue.

This page is significantly influenced by Robert Kegan’s psychological framework. He describes stages of personal development over a lifetime, numbered 0 to 5. Adults may progress from the communal, pre-systematic mode (stage 3), through the rational, systematic mode (stage 4), to the fluid, meta-systematic mode (stage 5). I’ve labeled the seven snapshots 4, 4.2, 4.4, and so on. The decimal notation is not quantitative; it’s an artificial expository device for pointing out the gradual nature of the transition.3 These steps are not sudden jumps, but arbitrary markers, spaced about a year apart, in a continuous process. By breaking the long transition down into pieces, I hope to make it, and the endpoint, seem feasible and understandable, and not inscrutable magic.

Each snapshot shows Prithi participating in more-or-less the same event. Retelling that story over and over lets you compare her reactions at different points in her development, apples-to-apples, as it were. Still, the details change as she reacts differently. Meanwhile, in the background, you can imagine her company growing from ten people to a thousand over the course of the page, and the several years of development it represents.

There may come a point, part way through this page, where it stops making sense and starts to sound abstract or implausible. If its overall trajectory seems attractive, you might consider the possibility that this point represents your own growth edge. Reflection around that step might be particularly helpful.

4. A system for relating

Yesterday the big sale Carlos was working on fell through. It was supposed to be a three-year contract with a giant multinational. They backed out at the last minute and signed with a bigger, older competitor of ours, who are “more stable,” although their products are inferior. It turns out our prospective client was sending all our confidential technical proposals to the other guys, and stringing us along to get our ideas for free. Carlos was so mad at them, and burned some bridges. But he also blamed himself for not figuring it out earlier.

We left the office right after; he told me he was going to start yelling at his staff if he didn’t get away right then. We did a long walk along the Embarcadero and then way the heck out to the Presidio. I was mostly just listening. That’s the best way to handle it when someone’s upset. Don’t try to fix things or suggest solutions or tell them they did or didn’t do the best they could. Anyway, when we founded the company, we agreed that sales was his thing. I stay out of it, except when he needs me there to explain tech stuff to a prospective customer.

He kept asking me if I’d be as angry as he was. I was pretty sure I wouldn’t have been. The outcome isn’t fatal, and who would have thought such a respected company would be so flagrantly unethical? But I couldn’t tell him that, because it would have been ignoring how upset he was. And he and I are really different, although we work well together. That’s a lot of why we work well together! My explaining how I would have felt would be irrelevant to how he did feel. It would be like making him wrong for feeling that way. Or at least interfering in his work to make sense of his own emotional process. Instead I wanted to just let him know that I understood how he felt, without judging it.

Prithi’s monologue might at first sound like the pre-systematic communal mode, because it is about feelings and relationships, which are central for that mode. But stages are defined by dynamics: by how you process meanings, not by which meanings you process. Relationships and emotions are always important. The question is whether you treat them pre-systematically, systematically, or meta-systematically. So understanding what makes this monologue systematic, not communal, will bring us to the starting point of the journey from systematicity to meta-systematicity.

In the communal mode, differences in feelings, opinions, and goals are inherently a problem for a relationship. In that mode, Prithi might try to persuade Carlos that he should feel about his failure the same way she did. Or she might think she should feel about it the way he did. Or she might feel like she wasn’t being a loyal cofounder because her views weren’t aligned with his. Startup cofounders in the communal mode try to sweep conflicts under the rug, which stores up trouble for later.

Prithi worked hard to understand Carlos’ feelings, but it didn’t bother her not to feel the same way. In the communal mode, you can take another person’s point of view; but in doing so, you lose your own perspective, because feelings are supposed to be shared. Prithi maintains her own feelings at the same time she recognizes his. For the systematic mode, emotional differences are not inherently a problem, because the relationship itself is formally structured, not just a bundle of of shared feelings.

Prithi exercises her systematic theory of how emotions should be processed in relationships, which includes not interfering with Carlos’ process. It’s not that Prithi is oblivious to Carlos’ feelings, or unsympathetic. She cares about them as a matter of genuine relational concern, as well as pragmatically and ethically. However, his experience is his experience; she’s “just visiting” his way of being. According to her conceptual framework, a successful outcome of this interaction will result in better understanding, but it will leave both of their ways of feeling and relating untouched.

Prithi defines herself as a rational system, interfacing with another rational system, according to a well-defined API. We call that API “professionalism.”4 It includes understanding that people have different roles and responsibilities in an organization, and different ways of making sense of feelings, people, and work; and that we work best together by respecting and allowing those differences. Prithi understands not just that Carlos has different feelings about the deal falling through. At stage three one can recognize that. At stage four, she understands the larger picture that his conceptual structure for success and failure and business ethics, which generates the feelings, is different. She understands that he too has a stage four worldview; he generally subordinates his feelings and personal relationships to his principles and his organizational role. For example, by not taking out his frustration on his staff!5

But… she’s being a bit rigid in her professionalism, in her non-intervention, in her insistence on sticking to her specific theory of how to do the relationship when he’s asking for something different. Prithi takes herself as the author of her personal system, and so—ironically—gets unconsciously defined by it. Synchronizing feelings for the sake of relationship harmony seems like regression to stage three for her. At the level of meaning, her problem is to figure out how to maintain her policy of keeping feelings to herself in the face of his repeated request that she violate it. His request is reasonable, but at stage four it does not occur to her to question the model.

She was, therefore, responding to Carlos not as a whole person, but in terms of an abstract principle, and in terms of their formal roles in the company. What makes this stage four is not this particular principle (a meaning). It’s the dynamics. The way she’s holding the meaning, as a fixed fact (“the best way”), has limited her space of possible actions unnecessarily.

Moving on toward the fluid mode, Prithi will gain greater freedom and flexibility by relativizing roles, principles, and systematic frameworks. But at this stage of development, she cannot yet see that she might operate effectively outside her own system, because she is subject to it as an ideological commitment. She would understand “operating outside the system” as just a reversion to unprincipled communalism. If meta-systematic fluidity were explained to her, she’d misunderstand it as an incoherent attempt to compromise between systematicity and the communal mode.

4.2. A glimpse of meta-systematicity

Carlos kept asking me how I would feel if I’d screwed up like that. And I kept turning the conversation back to how he felt, because I was pretty sure I wouldn’t have been nearly as upset as he was. I kept waffling and evading the question, because I didn’t want to invalidate his feelings. I thought they were reasonable, even if I would have felt differently.

But afterwards, I woke up in the middle of the night, replaying our conversation. We’ve been friends for years, and of course we sometimes disagree, but I think we’ve done a good job of resolving conflicts by asking “What’s best for the business? What evidence do we have? How do we do a cost-benefit analysis here? Whose area of responsibility does this decision lie in?” But recently I’ve started to worry he feels I’m distancing him. We’re still getting along, but are we still friends? It’s not that I’m not taking his emotions seriously; but we don’t want to get that in the way of our work.

I wonder if, in that conversation, I might have closed off an opportunity for him to learn something. I mean not that I would be teaching him, but just explaining a different way of thinking about it. Maybe my attempt to honor his way of thinking was counterproductive? Or, not the honoring, but my evasiveness… Maybe the way I decided how to respond is not optimal. But I’m not sure what I could have said that would’ve changed how he thought, or whether it could help for him to hear “no I wouldn’t feel that way.” Maybe I need to read more about interpersonal dynamics.

Or, I guess, maybe I should somehow have been feeling for some more intuitive or creative way of working with his upset? You can get lost in psychobabble woo that way, though.

In this second snapshot, Prithi recognizes that her principled, rational, systematic ideas about how relationships should work are open to question. She has some ability to imagine a meta-systematic alternative, but mostly can’t yet act on it.

At stage four, Prithi thought she had a rational theory of relationships, but in fact it had her. Because she took it as a justified, true belief, she was that system of relating, as nearly as she could manage. In the first monologue, the conflict she described was between her principles and Carlos’ repeated request. Now she experiences a conflict between her principles and her growing recognition that they are not absolute, and are open to doubt. She has begun to recognize that she is subject to a system. This is a first step toward a self larger than any ideology, with the greater freedom and power of reflecting on and using multiple systems.

If you do not share Prithi’s particular theory of relationships, you may be feeling superior now. She is being a bit stubbornly stupid. But recall that specific meanings—such as this theory—are not the point. What matters are the dynamics of how she relates to it. Holding any other principle in an absolute way would equally imply that she is operating in the systematic mode (and being a bit stubbornly stupid). This stubbornness responds to the fear that the only alternative to rigid rational principles is an irresponsible, emotional relativism: the communal mode she has left firmly behind.

Confronted with evidence that a principle is not always right, the systematic mode attempts to fix the framework; or—failing that—to replace it with some other explicit truth. (“Maybe I need to read more about interpersonal dynamics.”) The content can change, but within stage four, the way it is held remains the same. It retains an unconscious commitment to the idea that some system must be correct.

In 4.2, the possibility of reflecting on “the way I decided” arises. Prithi begins, hesitantly, to expand beyond her personal system. She begins to have some perspective on it, some recognition of how it works, which allows her to question its costs and adequacy.

On the way in to stage four, you fear that you may not always be able to conform to your system. On the way out, you become dissatisfied with its reliable bureaucratic functioning, because it’s emotionally isolating, intellectually limiting, and wearyingly familiar. You may wake up in the middle of the night and wonder how your life became so superficial, mechanical, and stale.

The first step beyond systematicity is suspicion. That begins after seeing enough systems fail, and enough irreconcilable conflicts between pairs of systems that both seem more-or-less valid. Is it possible that loyalty to any system might be a mistake?

So Prithi now wonders if she could “feel for” some other way of being. The nebulous possibility of fluidity, a new dynamics of meaningness, appears on the horizon, although she says she doesn’t know how or whether it could work. At this point, it is common to describe the imagined future way of being as “intuitive” or “creative.” This is not exactly wrong, but it expresses a sense of mere unknowing; of the possibility being too vague to describe. That can be frightening. Exiting stage four, as the personal system you identify with starts to lose its grip, it resists its loss of control. It makes progress toward five feel like regression toward stage three (“lost in psychobabble woo”).

Prithi also still understands the problem as one of how she would “work with” his upset. That is an individualist, stage four conception. Stage three tends toward emotional fusion, and stage four toward emotional self-containment. Stage five understands the self-other “boundary” as patterns of interaction that are real and necessary, but nebulous and permeable. Self and other actively co-construct, and ongoingly redefine, systems of meaning, including what it means to be inside, outside, or shared. But here Prithi can only distantly sense this possibility.

4.4. Learning meta-systematic skills

Carlos asked me several times how I would feel if I’d screwed up like that—and gotten screwed like that. At first I kept turning the conversation back to why he felt what he did. But eventually I realized I was being annoying and unhelpful. So instead I told him explicitly that it seemed like explaining that would distract from what he was feeling, and from helping him understand why.

And he said “But that’s not what I want from you! I know what I’m feeling, and why. What I want to know is whether there’s some better way I could be dealing with it. Admit it, I know you wouldn’t get so angry! Why not?”

So then we had a long talk about different ways he and I react when things go wrong, and how we make sense of success and failure, and about ethics in business. In the end he said that he realized part of his upset came from having assumptions violated that are maybe excessively rigid and unrealistic. And I realized that saying what I would think or feel or do in his shoes could be helpful, even though sales isn’t my area of expertise. Just listening non-judgmentally isn’t always the right approach. So I learned something, too. I think I have a better understanding of how to deal with other people’s upset.

Also, I think it’s improved our relationship; there’s a new sense that we can take input from each other in ways that go beyond just our professional roles. Our having to interface professionally with each other, while carrying all the responsibilities in this business—which are getting intense in new ways, as it’s grown—doesn’t mean we aren’t friends. Again.

In this snapshot, Prithi opens to the possibility of co-creating the meaning of the event. By explaining her evasiveness to Carlos, and then dropping it, she allows for a revision of her theory of relationships in the light of his different understanding. These are meta-systematic dynamics: they subordinate the systematic ideology to the process of ontological remodeling. This is the qualitative, collaborative transformation of one’s self, eventually allowing its meaning to remain inseparably nebulous and patterned.

However, at this point Prithi takes the possibility as one of theory revision to improve a systematic self—a stage four dynamic. Stage four is all about holding true beliefs, and about methods for successful problem-solving. Stage five takes those for granted; its dynamics focus instead on ontological transformations of nebulous circumstances.

In their conversation, both Carlos and Prithi discovered bugs in the operation of their personal administrations, and installed fixes they designed in consultation. They have co-created meanings, but each will own them individually, and as new but fixed commitments. In 4.8, we’ll see meaning scintillating in their interaction, jointly owned and constantly re-forming. At 4.4, even if revisions require significant re-architecting of their personal processes, the improvements will merely return each to a state of smooth, systematic individual functioning. That is the aim of stage four self-management.

So, overall, 4.4 is defined as having developed some skill in meta-systematicity, but not deploying it consistently. Also, it understands meta-systematic operations mainly from within a stage four conceptual framework, as ways of enhancing a system. In the cognitive domain, 4.4 is excitement about learning meta-rational skills, but taking them as new ways of accomplishing the goals of rationality, such as discovering truths and evaluating principles.6 This is not wrong, but it is a limited view.

4.5. The chasm of nihilism

I kept trying to just let Carlos go on about his feelings—because what’s it matter what I do or don’t feel? People have feelings, and they don’t mean anything. You just have to get on with the job. Different people have different feelings; so what?

Anyway, I was pretty sure that talking about it would just mean unloading my negativity on him, and what good is that. But he insisted, so I decided to just give it to him straight. I’m tired of pretending.

“In the bigger picture,” I said, “you could say we all just evolved to maximize our personal advantage, so you can’t really expect people to do anything else. You could say those guys behaved ‘unethically,’ but basically that’s just a social convention. Well, of course they were doing what was best for them! We happened to be collateral damage, but that’s just the luck of the draw.

“They’re in business to make money; we’re in business to make money. We’re past the point of money meaning anything; it’s just more of the same. We could walk away from the company now and be done. Why not? What do you have to prove? If you want to keep score by counting dollars and deals, that’s fine, but you might as well be playing a video game. It’s empty. If you run your life by your sales numbers, it’s going to be an endless emotional rollercoaster, and about nothing.”

Carlos didn’t like any of that. As I expected. He went on about “ethics” and “our mission” and stuff. It’s all BS, and he must know that at some level, but he’s still pretty idealistic.

Well, we don’t think about things the same way, but it doesn’t matter. We can still work together, I guess. It’s tiresome, though.

Something appears to have gone badly wrong here.

Crossing the bridge from four to five, it is possible to fall into a gap of nihilism. That is called “4.5” in the developmental literature. It is probably not a necessary step, although it is quite common, particularly for tech folks.

In my cheery depictions of 4.2 and 4.4, Prithi is drawn forward by a gradually clearer view of stage five: a better way of being. At 4.5, we see how she has also been shoved forward by an increasingly clear understanding of the defects of stage four.

In the systematic mode, you must semi-consciously blind yourself to anything that contradicts the system. In a group, everyone colludes to pretend things are going as they should when they aren’t, and to hope everything will work out somehow. When you see through this, it’s nauseating and infuriating. How could all those people around me be so stupid?

Systems offer false promises of meaning. When you’ve seen those fail enough times, you resolve never to get fooled again. For instance, you recognize that corporate “mission statements” are sanctimonious kitsch, designed to dupe dullards. The lie becomes increasingly offensive, and it’s hard not to attack it.

In relationships, rigorously containing your emotions, managing the emotions of less systematic people, and carefully channeling all interactions productively—skills that were triumphant accomplishments of stage four—come to feel like endless scutwork. Holding everything together, emotionally and practically, for people who are too childish to hold themselves together becomes claustrophobic and oppressive. Part of you wants to just say what you feel, but you know precisely why that’s a bad idea.

You recognize that that all ethical systems are shams. None of them can give the absolute grounding for moral judgements that they claim. They all sometimes give awful advice. Anyway, mostly there is no right or wrong; it’s all gray areas. At 4.5, you may adopt explicit amoralism.

From 4 through 4.4, systems give definite answers to questions of meaning: you are justified, by principles, in caring about the things you do.7 Fear of losing that caring is a major reason for holding back from development, for getting stuck. That fear is realistic! At 4.5, “what is meaningful?” becomes an oppressive existential quandry, or has a definite negative answer. You may convince yourself that everything is completely meaningless: the stance of nihilism.

Prithi sounds contemptuous of Carlos’ assertions of meaning, because at 4.5 they sound like naive idealism, which she has put firmly and permanently behind her. (The word “just” often serves to deny or minimize meaningfulness. Notice how often Prithi used it here.)

When you lose all faith in systematicity, but can see no workable replacement, you may fall into crippling dysfunction. I know people who, at this point, became completely non-functional for several years. Feeling that you have lost the capacity for confident, competent self-administration can render you practically catatonic.

Prithi has a milder case. She is still functioning as a startup CTO, although she may often feel that she’s running on empty (“you just have to get on with the job”) and her staff may now find her bitterness frightening and demotivating. She is cynical, but not (yet) despairing. In the language of Meaningness, she displays mainly materialism, the stance which denies “higher purposes” but admits “mundane” ones—“personal advantage,” as Prithi says. Sometimes she goes a step further, into Lite nihilism: the business is “past the point of meaning anything.”

4.5 is a profound change in the dynamics of meaningness, based on an ontological misunderstanding, and usually supported by extensive (although fallacious) intellectual reasoning. It is similar in feel to depression, and they often go together, but nihilism is not primarily a mood disorder. Mistaking it for generic depression can actively impede moving beyond it.

I said “something appears to have gone badly wrong.” However, there is much that is right in 4.5 nihilism. It is a genuine growth step forward from 4.4 because its recognition of the limits of systems, and the defects of systematicity, are accurate. It’s true that “missions” are usually self-aggrandizing or cynical propaganda (although genuine purpose is possible). It’s true that ethical systems are all fallacious and sometimes harmful (although accurate ethical judgement is possible).

Post-systematic Prithi does have a “bigger picture.” In the first two sentences of her lecture to Carlos, “you could say… you could say,” she points out the relevance of two ethical views, without holding out either as an absolute. She has relativized not just a particular system, but systematicity itself, so she is no longer subject to any ideology. What she cannot yet manage is to coordinate multiple ethical systems, in context, to make a wise meta-systematic judgement.

In her relationship with Carlos, she is newly brave and honest. This was probably an exceptionally bad time to start, but a new sort of intimacy has become possible. She’s finally willing to break out of her system, to expose her feelings and her way of making sense. (In 4.4, she discussed her feelings conceptually but did not allow them to break through into her way of talking.) She’s suspending her rigid self-containment for a moment, taking a risk, and “giving it to him straight.” The “tiresome” burden of pretense no longer seems worth the effort.

Post-systematic nihilism results when you finally give up on systematicity, but find nothing better to replace it with. If you have developed meta-systematic skill before letting go of systematicity, and can see the way forward, there is no need for nihilistic hostility, depression, or anxiety.

Let’s rewind to 4.4, imagine that Prithi was able to mostly skip 4.5, and head cheerily onward into 4.6.

4.6. Pulling away from systematic limits

In this snapshot, as in 4.4, Prithi displays a mixture of systematic and meta-systematic understanding. What’s different is that the overall dynamics are now fluid (although carrying significant systematic baggage), whereas 4.4 had systematic dynamics overall (although Prithi was growing significant meta-systematic skills). The balance has shifted toward fluidity.

At first I evaded Carlos’ question. My policy, at least as a default, is mostly just to listen, because when people are really upset, they need first to talk, and to feel that they are being understood. Then I can play back what they’ve said, maybe in more precise language than they’d managed, if I’m following their process accurately. That might help them see more clearly how they are thinking about it. Or they may say “no, it’s not like that,” and I’ll get a better picture myself. I don’t interfere with their process unless it looks like someone’s headed off a cliff. It’s much better for them to work through it themselves. And Carlos wasn’t heading off a cliff!

But Carlos didn’t want to talk about how angry he was; he wanted to talk about why I wouldn’t be angry. So for a couple minutes I felt kind of stuck. That’s not how this sort of conversation is supposed to go! I’m a pretty good people manager, for a techie, and I know how this works. I’m not going to let go of that so easily.

The residual fourishness in this monologue is that Prithi has a set system for addressing a particular sort of situation (“my policy”). On the other hand, she recognizes that it admits exceptions. It is only “a default,” although she “doesn’t let go of it easily.” She would like to restrict the exceptions to specified criteria (“going off a cliff”), thereby including the limits of the system within the system. But, in this case, that doesn’t make sense, which for a while leaves her feeling “stuck”—because she’s clinging to her old, systematic way of being.

But Carlos is incredibly smart in his own way, and I trust him even when he does something I don’t really understand. So I started explaining how I would have felt, and that opened out into a broad discussion about what success and failure mean.

Here she breaks free of her stuckness, and suspends her policy, operating outside any criterion. That’s fiveish.8

And about why we are even doing this! I mean, when we started, obviously we wanted to be successful, we wanted to make buckets of money, we wanted to make cool stuff, but I don’t think we had really thought that through. Now we’ve done all those things. Do we check off “found a startup” in our task management apps and move on to the next thing? “Success” means something different for us now than it did then. It’s not really clear what. Just setting the same bar higher—more customers, more employees, bigger valuation—that’s not going to be satisfying in the long run.

I’ve been struggling with this myself for the past year or so. I’ve wondered if I’ve gone as far with CTOing as I want to. I hadn’t admitted this to Carlos, but I’ve toyed with the idea of leaving the company to take on something quite different. Maybe I should start an Effective Altruism organization instead, for example. But in the end, whatever your project is, it will succeed, or fail, and you’re just back in the same place. When you don’t know what you are capable of, doing anything difficult is really exciting. But when you’re confident you can succeed at most things, you have to ask “do this? do that? do what? why do anything?”

This echoes Prithi’s nihilistic doubts at 4.5. She recognizes, rightly, that any definite positive answer to “what is meaningful?” will ultimately fail: “you’re just back in the same place.” But at 4.6, Prithi glimpses an alternative to both the positive certainty of stage 4 and the negative certainty of 4.5 nihilism. Meaningness can become a permanently open question. The possibility of groundless caring begins to emerge.

However, another part of her still has the sense that finding her true purpose is a problem that could somehow, in principle, be solved. “It’s not really clear what” suggests that the issue is epistemological: there is a correct answer, a truth, but she doesn’t know it. The fluid mode recognizes the problem is ontological: purpose is real and vital, but necessarily nebulous.

So it turned out we talked as much about my feelings as about his! Carlos does think about things very differently from me, and he wound up intervening in my process at least as much as I was intervening in his. Although I think he was honestly mostly just trying to understand how I was thinking about it.

In retrospect, I wish we’d talked this through a year ago. I’d been stuck with some pretty fixed ideas about what we are doing and why, and they weren’t working anymore. The discussion has freed me up to reevaluate my role, and to talk about how my real attitude toward the company has kept changing, year after year. That was a huge relief. Admitting my doubts felt like a big risk, but it turns out he’s supportive of whatever I decide to do.

Prithi now values the reevaluation of ontological commitments more for the sake of allowing the ongoing unfolding of meaning than for any specific systemic improvement it may deliver. This is definitional for stage five.

Prithi attributes this move beyond systematicity to her trust in Carlos. It would be equally accurate to attribute it to her growing trust in her competence in navigating territory she “doesn’t really understand,” without any map to guide her. She’s excited not that she has discovered new truths, or that the interaction has improved her self-system, but that it’s that it’s freeing her from taking any system as a fixed truth.

She reflects meta-systematically on the limitations of her own mostly-transcended systematic self. Her default way of being is shifting into meta-systematic fluidity. The question of her role is no longer a problem to be solved, but an inherently open-ended process of transformation. Involving Carlos in defining the meaning of her work life is now a reality, and not just a matter of occasionally “taking input” from him. On the other hand, she still partly conceives of this as an individual process with a definite end-point. She speaks of “his process” versus “my process,” and says she will “reevaluate” her role, and “decide what to do.”

A possible misunderstanding here is that the shift from systematicity to fluidity is one from goal orientation to process orientation. In fact, you can prioritize either goals or processes (or, ideally, both) at any developmental stage. What differs is the way you relate to them.

4.8. Mostly meta-systematic

Why wouldn’t I have been angry? For a couple minutes, I deflected Carlos’ question, saying I wanted to give him space to work things out himself. But he called me on that. When we discussed my reluctance, I realized that, although that was genuinely one motivation, I was also hiding something, partly even from myself.

As we talked it through, I had to admit that, in Carlos’ position, I really would have been angry. I could be calm about it only because I considered sales not my problem. That’s been the fundamental division of responsibility in the company from the start. “I’m in charge of tech, and I deliver.” It was often a gargantuan effort, and I thought I had to be in control of everything to make it happen. But in enterprise sales… you can’t be in control when you’re trying to make a deal with a company a hundred times larger than you. Meeting engineering targets is tough, but he’s had to deal with way more uncertainty than I have.

I knew that, but I’d always implicitly considered it his problem—because I didn’t want to deal with it. On reflection, I knew I’d often have been terrified. And, yes I’d be furious about their double-crossing us. I mean, I was furious, once I admitted that. Although it was also moderated by my taking success and failure more analytically than Carlos does.

When we began, it was just the two of us, so “CTO” was almost a joke. It just meant I wrote all the code, while he hustled for money. We both also did whatever was necessary. I did the bookkeeping and assembled the flat-pack furniture and negotiated our first office space lease. But then we hired people to do all those things. And once we had half a dozen engineers, “CTO” really meant system architect: I made the design decisions. We grew some more, and “CTO” came to mean hiring and coordinating our system architects, who understood technical issues I couldn’t keep track of anymore. That was hard, letting go of complete control of the technology! And recruiting and line management is not what I enjoy, or am especially good at. It’s really the job of a VP of engineering. So we hired one of those, and now—at last—I’m a real CTO.9

My job is to communicate an inspiring technical vision, inside and outside the company: to our staff, enterprise customers, partners, investors, the IT media, and the professional community.

And… (deep breath) that’s a sales role.

In discussion, together we realized that, although officially we are co-CEOs, I was avoiding taking full responsibility for the company’s success overall. And by doing that, I was artificially limiting my effectiveness—and maybe the company’s effectiveness.

It was comfortable to think of myself as a geek, just in charge of the tech side of the company. But that hasn’t been completely true ever, not from the beginning, and particularly not since I moved into the real CTO role. So… I’m going to rethink “co-CEO,” and take it as meaning not just that Carlos and I make the big decisions together, equally—as we always have—but that the whole company is my job. And his job. Even if we emphasize different aspects.

I’m feeling a bit of vertigo. It’s going to take some work to keep reminding myself of how it feels to expand that way; to not fall back into just being Alpha Geek. But it’s also exciting looking forward to what I’m becoming next!

Prithi increasingly lives as the dynamic space within which diverse systems operate, interact, and evolve, rather than as a single static structure of principles for action.

That dynamic space is not enclosed by her skull, nor limited to her sphere of responsibility. It is co-defined by Carlos, and to varying extents by everyone she interacts with. It includes “our staff, enterprise customers, partners, investors, the IT media, and the professional community.” This is not communal-mode merging; she’s still perfectly clear that she and Carlos are dissimilar people. (And it’s certainly not the monist fantasy of “becoming one with everything.”) It’s an understanding that meaning plays out in interaction, not in her head. She has not ceded territory; she has expanded by recognizing that she contributes to, but does not need to individually own, the space.

The discovery of Prithi’s self-protective strategy was not the result of personal, psychological introspection; it was collaborative. And what she and Carlos uncovered was not primarily mental contents, but something she was doing—revealed by observing what she was not doing.

The central event in this monologue is this insight that Prithi has been hiding an internal conflict from herself. Such insights can occur at any developmental stage. Let’s look at meanings of such events, and dynamics that lead to conflicts and to regressive or insightful resolutions.

Understanding how each mode resolves conflicts of meaning may be the best way of understanding how the mode operates. Likewise, understanding how each fails to resolve conflicts may be the best way of understanding how you move to the next one.

In stage three, conflicts between desires are subordinated to the maintenance of a relationship. In stage four, conflicts between relationships are subordinated to the maintenance of a system. In stage five, conflicts between systems are subordinated to the maintenance of an open space for meanings to interact.

In stage three, you may hide an emotional conflict from yourself because it can’t be accommodated in the context of the relationships it arises in. For example, you may feel torn between loyalty to your boss, who has made an unpopular decision, and loyalty to your team, which is conspiring behind her back. Such tensions surface during the three-to-four transition. The communal mode could only resolve this by denying the meaningfulness of one of the two relationships completely, leaving the underlying emotional conflict unaddressed. That would be a regression. Alternatively, the event could drive the insight that you need organizing principles to distinguish the different responsibilities you have in different sorts of relationships; how each sort is limited; and how to relate different relationships to each other. In the work context, professionalism is the system that provides that structure. Its dynamics expand from the personal relationships in a social community to include the systemic role relationships in a work group—regardless of the personalities involved.

In stage four, you may hide a role conflict from yourself because it can’t be accommodated in the context of the system it arises in. This may be a conflict between two roles (Prithi is both CTO and co-CEO) or between formal role boundaries and reality (Prithi is officially responsible only for tech, but cares about everything in the company and has had a de facto sales role). Such tensions of definition surface during the four-to-five transition. The systematic mode could only resolve this by denying one side of the conflict, by insisting that Prithi’s responsibility is limited to a single formal job description. She would have to blind herself to her competent contribution to aspects of the business that “aren’t her problem.” That would be a regression. Alternatively, the event could drive the insight that systems are always artificially limited, and she needs meta-systematic skills to relate systems to each other and to reality. Snapshot 5 and the Epilogue will show how that works in startup leadership.

In both transitions, you need many small, concrete insights before you get the hang of the pattern and move to the next stage. That’s why it takes years.10

At 4.8, Prithi is most of the way there, and so resolves this conflict meta-systematically fairly easily. What she and Carlos uncover is her pattern of protecting herself from fear of nebulosity (the amorphous uncertainty of enterprise sales) by applying a limited, formal self-definition. As that self-system reaches its limit and breaks down, she lets go of the temptation to cling, and instead takes it as an opportunity for expanding her view.

Prithi now understands herself as having multiple self-systems, all valid. So there is the Prithi who protected herself by limiting her responsibility to what she could control—which is totally valid. When she recognizes it, Prithi does not chastise or abandon that part of herself. She just will no longer be ruled by that definition—or any other. There is programmer-Prithi, who thinks the most important thing is to get the technical decisions right. And that’s true! And there’s engineering-manager-Prithi, who thinks the most important thing is to make sure the development team operates effectively. And that’s true! And there’s CTO-Prithi, who thinks the most important thing is to create and communicate a vision of the future—and that’s true! So long as she is open to all these possibilities, no conflict is necessary. Meta-systematicity encourages respect among systems, even when they contradict irresolvably.

However, Prithi here is not yet quite at stage five. There’s still a trace of over-emphasis on an individual viewpoint: “what I’m becoming next.” At stage five, the space of meaning is not personal. Here at 4.8, holding that space open is still challenging (“a bit of vertigo”; “will take some work”).

In a context in which one of her systematic selves is expert, it will naturally try to take control, and she will need to remind herself not to identify with it, even temporarily. Meta-systematicity goes beyond knowing you are several selves you can switch between. Fluidity does not attempt to construct a meta-system for choosing what system to apply when. Nor is any system either alien (unambiguously outside) or a possession (unambiguously inside).

5. A deliberately developmental organization

Carlos and Prithi
Some years older and wiser
Image courtesy rawpixel

Yesterday the strategic partnership arrangement we’d been working on fell through. All along, they were intending to do a deal with another company, and were pretending to negotiate a partnership as a way of getting a deeper look into our technology development roadmap.

We took our usual long walk out to the Presidio, to decompress and talk it over. We were thoroughly annoyed, which we thought was funny. After all these years, we still get ticked off when someone pulls this sort of nonsense? It’s also funny that Carlos still takes it more obviously emotionally, whereas I at least pretend to be detached and analytical in my anger. Hot and cold.

Apparent interpersonal conflict may be as intense at stage five as at any other. However, it occurs within the fluid open space where meanings interact—just as conflicts between one’s own selves do. Stage five identifies with the space, not with one’s personality or chosen principles. That doesn’t negate personal differences (“hot and cold”), nor partiality, nor principles, nor strong emotions. It does give freedom to watch conflicting meanings interact across the boundaries of persons, organizations, or tribes, without an automatic compulsion to defend some particular one. Sometimes one can learn even from people who are being hostile or unethical.

Recognizing contradictions without needing to resolve them can be entertainingly absurd. Prithi’s amusement at her own pretense in hiding her anger, and Carlos and Prithi laughing at themselves in conflict with others, is a sign of their being meta to their own self-systems. Their combination of anger and humor also shows their being larger than identifying with their own company. That wouldn’t, for instance, stop them taking the other company to court if warranted, though.

Nowadays we’re mostly annoyed with the default culture of business that says selfish game-playing is normal and acceptable—more than being mad at specific people or companies. But we’ve realized selfishness is an intrinsic aspect of being human. You can’t remove it; and anyway, it’s not inherently bad. Wanting more is part of why we work and create, as well as motivating harmful or ethically sketchy behavior.11 So we try to work out structures that minimize its negative consequences. Nothing works perfectly, but active measures have helped. We’ve had to come up with new approaches every couple years, as the company has grown.

This final snapshot shifts the focus from Prithi’s relationship with Carlos to their joint facilitation of effective relationships among others. That is, from personal to organizational psychology. Prithi and Carlos are becoming the creative, reflective space within which others evolve.

Each major expansion has meant we had to put in place qualitatively different organizational modes. It’s not just formal structures; the way everyone relates to each other has to change, too. That’s been difficult, for us and everyone.

The company’s gotten big enough that Carlos and I can’t know what’s actually going on inside. We still try, but mostly we rely on our group leaders. And the natural tendency in their position is to try to grab and hold territory, to expand their division, so they look important. Then you get departments working against each other, which is how big companies get slow and stupid and full of politics that makes everyone miserable.

So we try to redirect that drive, guiding our leaders away from empire-building. We reward them for letting go of definitions, for seeing bigger pictures, not for gaining territory.

When I realized that CTOing is partly a sales role, I had to do that myself; to rethink what I am. I took several sales training courses—and got Carlos to coach me, too. That probably made me a better CTO, but it also changed the way I feel about and relate to people in general. Sales, for us, is about creating genuine connections, not tricking people into buying things they don’t need. Skill in connecting is helpful outside work too.

Prithi has come to take her own fluidity for granted. It is a reliable fact that, in the face of difficulties, she will act out of a realm of unspecifiable possibility, not ruled by any systematic theory of business management.

When you manage a business whose details you can’t know, the temptation is to think of it as a big machine, or a manufacturing plant. Then you try to optimize output by moving boxes around in the org chart, and your actions are all “move 27 engineers from department A to department B,” even though you have no idea what they do. They are engineers, right? and the spreadsheet says we need to increase output in B. You think people are their job descriptions. Like machines on the factory floor that have defined inputs and outputs. Then they’re forced to pretend to be their job descriptions, and that’s a disaster.

A system of formal responsibilities is a critically important tool as you go from a hundred people to a thousand; but you also have to realize it’s just a representation. An org chart is not reality, and you don’t change reality by changing the representation. You have to keep asking “how does this representation relate to the nuts-and-bolts reality of how these particular people work together?”

Prithi here expresses a core concern of meta-rationality, which understands relationships between representations and reality not as truths but as tools. This is meta-systematicity in its cognitive manifestation.12

We can’t afford to have anyone limited by a job description. Like, when we started, I didn’t know how to negotiate an office lease, but it had to be done, so I read some stuff and then just plunged in. Realistically, nobody can do everything. When being a regional comptroller, you need to know the GAAP rules for amortizing development costs, and when being a front-end developer, you need to know the Javascript rules for type coercion. But in principle you should be willing to learn either or both, if you had to. Beyond that, there’s learning to think like an accountant or a programmer. Or, more realistically: for an accountant to think like a facilities manager or market strategist; and for a programmer to think like a product manager or UX designer.

So we aim to develop and reward “fluid competence,” more than excellence in a specific role. We encourage an attitude of “OK, this needs to be done, I can probably do it”—combined with wanting to actually figure out how to do it, not faking it or going through the motions or trying to stay safe by doing it by the book. And not covering up when you screw up! We praise and reward people for screwing up on hard things, if they are open about it.

Learning technical skills in very different fields is one of the best strategies for developing meta-systematicity—if you recognize that they imply different ways of thinking and feeling and acting, not just different lists of formal rules to master.

Last year, we introduced ongoing, mandatory interpersonal skills training for all employees. We hoped it would particularly help our more technical people broaden their competence. I knew a lot of the engineers would hate the idea. Some said the company was turning into a cult, and we lost a few of our best people. It was a calculated risk. Most stayed, and some say the training has radically improved their lives—outside work as well in it. Others resent it. Qualitatively, it seems to be a big plus overall. It probably shows up in our numbers too, but it’s not like we could have a control group. Maybe we’d be doing even better without it.

Our approach isn’t The Right Way. It’s one way, which is working, so far. It’s given us a unique reputation in the industry. We can’t be “a great place to work” for everyone. Many people prefer an environment where they can get on with doing the job they were hired for, which they know they can do well. I totally get that! My job often terrifies me. It would be way more comfortable to write code all day, or even to be a conventional CEO. But the unfolding world keeps enticing me into the unknown. “OK, this looks like the next thing, let’s figure out how to do it.”

Just as Prithi’s specific theory of relationships at stage four was not definitional of stage four, her specific approach to management here is not definitional of stage five. As she says, interpersonal training and rewarding broad competence are not The Right Way. What is fiveish is the attitude of reasoned experimental curiosity, not aiming for any final conclusion or achievement, but for ongoing responsive fluidity.

In 4.6, Prithi showed a wistful longing for some ultimate principle that would let her decide what to do—whether to leave the company for another project—even as she recognized that no such principle can exist. Here she’s left that behind without a trace. Purpose is an unrolling dynamic, a collaborative improvisational dance of self and world, continually revealing new openings and obstacles and their meanings. Stage five means letting go of eternalistic promises of certainty, understanding, and control, in favor of appreciating the inseparability of nebulosity and pattern.

Looking back over our past decade, the most striking thing is not how Carlos and I have grown the business, but how the business has kept growing us. We still have the capacity to surprise each other.


The story of Prithi and Carlos is about meta-systematicity in a two-person relationship. However, situating that relationship in organizational leadership allows me to switch topics here.

Meta-systematicity in organizations

Management theory is the domain where meta-systematicity is most widely appreciated, discussed, and understood. I plan to write about this in the “Fluid society” chapter of Sailing the Seas of Meaningness. That chapter will not be ready for some time, so I’ll make some preliminary observations here now, out of place.

Entrepreneurship is inherently meta-systematic.13 Entrepreneurs create companies, which are new systems. Especially in the tech industry, the products and services they provide are often also novel systems. Rapid growth requires constant reorganization. System transformation is the essence of fluidity, and usually the basis for successful entrepreneurship. (Although innovating by rote formula, like “Uber, but for X,” is much easier, and you may get lucky with it.)

It’s a commonplace now that all organizations operate in volatile, uncertain, complex, ambiguous (VUCA) circumstances. “VUCA” is business-ese for nebulosity, which the systematic mode attempts to deny or control out of existence. There’s widespread understanding that this is impossible, and that meta-systematic approaches are a better alternative. There is no solid ground; it’s a whitewater world, so organizations have to be fluid too.

A simplified history of the evolution of management practice may help understand this.

Traditional theories assumed workers were pre-systematic. The job of management was to provide a systematic framework within which employees worked, prototypically in manufacturing. The company as a whole could operate systematically, despite the pre-systematicity of most employees. (We could call this the “three in four” model.) Systematic executives could then optimize operations using rational, often mathematical, methods.

As postindustrial economies shifted to “knowledge work,” management practice recognized that individual employees may work systematically, and so can and must be given much greater autonomy to exercise their specialized expertise. At first this left the company’s structure unchanged: systematic employees worked in a systematic framework optimized by systematic executives. (“Four in four.”) They were understood as having specific technical, professional skills, which amount to the ability to act in accord with a particular rational system (such as GAAP or Javascript).

But as VUCA intensified, companies that had dominated their industries with relentless rational optimization and technical excellence suddenly failed when market conditions changed out from under them. Their systematic approach used fixed analysis methods that became irrelevant or inappropriate in the changing environment. Some companies are still stuck in this “zombie era” of management.14

Leaders had to learn to rethink their systems increasingly often. “Adapting to changing market conditions” means not just introducing new products or entering new markets or changing market position. It means rethinking the fundamental mindset of the company, because otherwise you find yourself weighed down by “how we do things” and unable to react.

This culminates in meta-systematic fluidity: continuous ontological remodeling. This is now almost conventional wisdom, and considered best practice, in dynamic industries such as tech.15 Meta-systematic management is widely admired in theory, but putting it in practice is difficult, and not so common even in tech. It demands meta-systematicity in senior executives.

Even then, the assumption commonly remains that most employees can only function systematically. That means that the company’s overall operation is still systematic at any moment in time, albeit with leaders injecting frequent doses of structural change. (“Four in five.”)16

Some now recognize that executives alone cannot provide sufficient fluidity. They can’t have sufficiently intimate knowledge of the details of their employees’ work—which may be even more subject to VUCA than the company overall. Just as a previous generation had to learn to trust workers to function systematically without detailed supervision, managers now need to learn to trust workers to function meta-systematically. That is, management needs to delegate continuous reformation of parts of the company to the employees who understand that part. (“Five in five.”)

This might sound “nice” and democratic to workers, or terrifying for managers, but increasingly it’s just necessary. No one knows how to structure organizations in the whitewater world; there are no longer any standard principles that work reliably. You cannot see the world clearly through the lens of a system—any system. Effective organizational functioning has to be a collaborative improvisational dance with the environment, figuring it out together as you go along.

You can’t afford not to involve as many of your people as possible in that effort. Ideally, every employee should contribute to the continual redefinition what the company is and how it functions. Realistically, most can’t do that today. The capacity must be developed. Especially, systematic employees must develop meta-systematic competence, so they can go beyond formal professional expertise to respond rapidly and accurately to emerging business conditions, exercising judgement that goes beyond any set criteria.

Because meta-systematicity is rare, takes years to develop, and is not taught in school, companies have to train it in-house.17 This makes for what Robert Kegan and his collaborators call a deliberately developmental organization.18

This requires huge management effort, but appears to have correspondingly huge financial as well as human payoffs—in Kegan et al.’s case studies at least. The deliberately developmental approach is not easily put into practice, but there’s a growing enthusiasm for it and increasing bodies of theory and practical resources.19

This understanding naturally extends to the concept of a deliberately developmental society, in which a nation-scale culture explicitly recognizes the value of adult development, not just teaching specific facts and narrow skills. In 2019, critical political, educational, and economic systems are visibly crumbling. It’s urgent to bring more of the population to systematicity; and bringing some others to meta-systematicity is critical to enabling that.

The development of the cofounder relationship

Some venture capitalists say that startup success depends more on the cofounder relationship than any other factor. It can be their main reason for choosing to invest, or not, in a founding team. Conflicts between cofounders may be the most common reason for startup failure.20 So, many resources aim to help strengthen these relationships: blog posts, podcasts, coaching, and bootcamps.

Tech startup founders are invariably cognitively systematic, from their education and early-career work in either STEM or business. However, many may not have developed to systematicity in their emotional lives and relationships. Most cofounder relationship advice I’ve found is about moving from being driven by emotions and personal relationships (stage three) to professionalism (four):

This is probably exactly what many straight-out-of-school founders need. It’s too basic for someone who’s gained some maturity from several years work experience, perhaps in a team leadership role.

Putting coherent organizational systems in place is the main job of scaling a startup from tens of people to hundreds. This page starts from the prerequisite, stage four. That is already more mature than many startup founders. When investors say “it’s time to bring in professional management” or “adult supervision,” the point may be to force a professional, systematic mode of relationship on the executive team. Or, to bring in technical expertise in building systematic administration, which the founders lack. Much of the work in this scale-up phase can be done by applying off-the-shelf systematic patterns. You do need org charts.

But system-building won’t cut it for more than a few years in a VUCA environment. A medium-sized systematic company is a duckling sitting in open water, soon to be devoured by piranhas (smaller, faster competitors) or a hippopotamus (a much larger one). You had better grow your meta-systematic wing feathers fast.

Some consultants and coaches offer meta-systematicity training. They seem to market their services only to senior executives in larger companies. There’s not much information available on the web, and there seems to be little awareness in the startup community. I hope that will change soon. Since entrepreneurship is inherently meta-systematic, earlier training in personal and organizational fluidity should be valuable. And as more companies adopt the “five in five” model, it will be increasingly necessary throughout the organization.

What did I just read??

I mean, what even was this stuff? And where did it come from?

If you find this page interesting, I would suggest considering what parts of it are believable or useful and why. You may have to proceed meta-systematically…

I have no relevant academic credentials. And, while I have started, grown, and sold a small, successful tech company, I was a solo founder. What I’ve said about business is mostly not based on personal experience.

My understanding draws on many fields; developmental psychology was the biggest influence on this page.21 Kegan’s is one of several broadly similar theories of “postformal operations.”22 These theories inspire me not because they are well-grounded as science, but because they make sense of my experience, and what I know anecdotally about exceptional prowess in technical research and in organizational leadership. Some experimental work has been done to test postformal theories. I’m not necessarily qualified to judge this research, and I haven’t investigated it in depth, but I haven’t found the studies scientifically persuasive. (Particularly in view of the current replication crisis in psychology.) I write about Kegan’s version because it’s simple and fits my anecdata, not because it has the best experimental support.23

So. You will have to figure out for yourself how to evaluate what you have just read. What considerations would be relevant? What would it even mean for it to be accurate or useful?

Such investigation is the essence of meta-systematicity—because there are no predetermined criteria or methods, and no preexisting problem definition or conceptual framework to decide how to think about it. If you choose to proceed, you will ask how my story relates to reality, and to other conceptual systems you know—crossing a chasm of nebulosity without a net.

Future directions

Although this is possibly the most detailed and practical explanation yet given of the path to meta-systematicity, it still seems unhelpfully brief and abstract. To make it more concrete, I have begun writing a “fluidity workbook” full of exercises. If I had a spare six months, it would also be fun to expand this page into a business novel. My inspiration is Eliyahu Goldratt’s cult bestseller The Goal. The book would take Prithi, Carlos, and their company from stages three through five, illustrating each step with a new business situation that would plausibly prompt the next form of development.

Although I’m deeply interested in meta-systematic leadership, I may be more obviously qualified to teach meta-systematicity to individual technical contributors. I wrote a preliminary curriculum sketch, “What they don’t teach you at STEM school,” a couple years ago. I am currently writing an introductory textbook, In the Cells of the Eggplant.


I am grateful to Gary Basin, Sarah Constantin, Zach Obront, Malcolm Ocean, Rin’dzin Pamo, Taylor Pearson, Venkatesh Rao, Graham Rowe, and Anand Vemuri for their suggestions on draft versions, which led to extensive revision and improvement. Of course, they may disagree with some or all of this, and aren’t at fault for whatever’s wrong with it.

  1. 1.Imagine explaining your doubts to a holistic crystal healer. “There’s no good evidence this works. It seems highly improbable, given everything we know about crystals and healing. Several of your claims are clearly factually false. When you say you rely on ‘holistic intuition,’ you just mean you are making it up as you go along; you have no specific method.” If the healer has not yet learned to think rationally—systematically—they can only hear this as a string of insults, or delusional assertions of your personal or tribal superiority. Words like “evidence,” “probability,” “knowledge,” “truth,” and “method” do not mean the same thing to them as they do to you. These terms function only within a web of systematic understanding of what systematic understanding is. This chicken-and-egg difficulty is part of the reason only a minority can think, feel, or act systematically. Analogously, a systematic thinker can only hear an explanation of the limits of systematicity and the advantages of meta-systematicity as delusional assertions of the superiority of inscrutable nonsense. This incomprehension is one reason so few can think meta-systematically.
  2. 2.In writing this page, I’ve drawn on a discussion of the development of a marital relationship from systematicity to meta-systematicity in A Guide to the Subject-Object Interview, by Robert Kegan and his collaborators, who I’ll discuss repeatedly in this page. I’ve re-staged it: in order to re-present the material without violating copyright; because the tech industry context may be more interesting to my readers; and as an interesting exercise for me personally. The Guide is a technical manual; see below for better starting points for learning about the theory.
  3. 3.There are other similar theories of adult development with different numbers of stages. I’ve adopted Kegan’s numbering somewhat arbitrarily. Details of his theory are open to doubt (as I discuss briefly below). However, this page does not depend on most aspects of the framework. It relies only on the conceptual distinction between systematicity and meta-systematicity, which many other thinkers have pointed out, in varying terms. For more on his framework, see my “Developing ethical, social, and cognitive competence,” or his The Evolving Self.
  4. 4.An API (application programming interface) is a formal definition for how a program can work together with another one. The value of an API is that the separate developers of the two programs can coordinate without needing to know anything about how the other program works, or how the other development team works. Also, a program can work together with any other program that uses the same API; it doesn’t matter which. Professionalism is a set of standards of behavior that aim to save you from having to understand all the details of your co-workers’ personal and emotional lives. You can relate to people in terms of their formal roles, rather than their personality quirks and transient upsets. That is: systematically, rather than communally. In practice professionalism can never be fully achieved, nor should it be, but as a default it is efficient and reduces interpersonal stress—for those capable of it.
  5. 5.If Prithi were subordinate to Carlos, it would be highly unprofessional for him to insist on her giving an evaluation of his screw-up. But they are co-CEOs: a risky and unusual arrangement that can nevertheless work well, because it gives the CEO someone to talk to as an equal. Carlos was also unprofessional in “burning bridges” with a potential client. That’s part of what made this a screw-up. Everyone screws up sometimes; professionalism is a commitment, not an app you can just install and let run.
  6. 6.Misunderstanding meta-rationality as an extension of rationality prompts the demand “just show us how these meta-rational methods work, if they are so great!” The assumption is that meta-rationality must be just a specialized class of rational methods, which might be excitingly different from other ones, like encountering a new branch of mathematics or a new programming paradigm. If they are of value, they must have the same purpose as other rational methods, and will be incorporated in and subordinated to rationality. At 4.4, one may have genuinely learned to think in a new way, while still regarding it as a specialized type of expert theory-revision and problem-solving within a comfortably rationalist framework.
  7. 7.This is “eternalism,” in the language of Meaningness. Nihilism simply inverts eternalism. The opposite of a wrong idea is usually also a wrong idea, although it may be a necessary step toward a better idea.
  8. 8.It’s not threeish, because she’s not abandoning her principled theory in favor of emotional sharing or relationship maintenance. It’s usually a good theory, and she may often act with reference to it—but she will no longer be governed by its rules, even when she’s acting in accord with it.
  9. 9.The Manager’s Path: A Guide for Tech Leaders Navigating Growth and Change is a practical manual for the series of personal and organizational transformations from startup tech lead to CTO, by Camille Fournier, who has lived them.
  10. 10.I have described this years-long path of self-transformation, progressing through repeated insights into limiting self-definitions and hidden conflicts, in a very different style elsewhere, as “shadow eating.”
  11. 11.Fluid operation above and around systems is very different from communal-mode operation without a system. Fluidity fully recognizes and incorporates the effective functioning of systems, which the communal mode is blind to. A possible misunderstanding is that fluidity is amoral because it holds no absolute ethical principles. Since conflicts between fundamental principles are sometimes unavoidable, none can be absolute. However, fluid ethics hold principles in high regard, and may deploy them more effectively than systematic ethics, because fluidity has additional resources for resolving dilemmas.
  12. 12.See also Venkatesh Rao’s “The Amazing, Shrinking Org Chart,” on fluid corporations.
  13. 13.This goes double for venture capital. Startup founders have only to find a way to create a successful company, mostly through one-off improvisations. Venture capitalists have to understand general patterns of business success. It is tempting to rely on systematic rationality (metrics, criteria, theories) when evaluating startup investment opportunities. That’s what you learn in business school, but mostly it doesn’t work. The best investors seem to deploy meta-rationality instead.
  14. 14.Thanks to Venkatesh Rao for this term.
  15. 15.Some early theorists of meta-systematic management include John Seely Brown (the visionary leader of Xerox PARC), Chris Argyris (who worked with Donald Schön), and Bill Torbert (who worked with Robert Kegan).
  16. 16.These models are necessarily simplified abstractions. Any real organization will include people at all stages of development, and different sub-organizations will operate somewhat differently. And, developmental “stages” are themselves simplified abstractions, which are heuristically useful categories but not ultimately true as ontology.
  17. 17.Ben Horowitz makes an inspiring case for in-house training, led by senior managers, in The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers.
  18. 18.An Everyone Culture: Becoming a Deliberately Developmental Organization. Schools increasingly fail to teach systematicity, so companies have to do that too. The book describes a developmental pipeline, implemented by a property management company, from pre-communal stage-two teenagers through to stage-five meta-systematicity. One advantage of training meta-systematicity in-house is that senior executives may be less panicked about delegating transformational authority if they’ve been intimately involved in training the necessary skills.
  19. 19.There is also a danger of turning meta-systematicity and/or the developmental approach into a management fad, replete with buzzphrase cliches, hype ungrounded in evidence, and superficial implementations that predictably fail.
  20. 20.A widely-quoted statistic is that 65% of startups fail as a result of cofounder conflict. I traced this back through the citation chain, and it’s erroneous. The original study sent a paper survey to venture capitalists in 1984 asking them how they related to their portfolio companies. (Michael Gorman and William A. Sahlman, “What do venture capitalists do?,” Journal of Business Venturing, 4:4 (July 1989), pp. 231-248.) The VCs cited “ineffective senior management” as the most important factor in 65% of troubled or failing companies. Since it is the responsibility of senior management to fix things, this is almost tautological, and the number in some sense should be 100%. In any case, the study says nothing about founders (as opposed to senior management in general), and nothing about interpersonal conflict (just “ineffectiveness”). Anecdotally, however, cofounder conflict is indeed a common way for startups to fail.
  21. 21.For those interested in business applications, Kegan et al.’s An Everyone Culture: Becoming a Deliberately Developmental Organization is the most recent publication. Also see Garrett McAuliffe’s “The Evolution of Professional Competence,” which reviews and synthesizes research on fluid management by Kegan, Schön, Torbert, and others. (Chapter 21 in Hoare’s Handbook of Adult Development and Learning.) For more general discussion of Kegan’s framework, I recommend The Evolving Self: Problem and Process in Human Development and In Over Our Heads: The Mental Demands of Modern Life, which I have summarized elsewhere.
  22. 22.For reviews, see Eeva Kallio’s “Integrative thinking is the key: An evaluation of current research into the development of thinking in adults,” Theory & Psychology, 21:6, pp. 785-801; or, in more depth, chapters 8-12 in Demick and Andreoletti’s Handbook of Adult Development. My thanks to Matthew Mezey for suggesting the latter.
  23. 23.Although I am inspired by the work of Kegan and his collaborators, I have no affiliation with them, and what I say may not present their ideas accurately. Conversely, I have reservations about some aspects of their work, and do not necessarily endorse everything they say.