When it comes to parenting, coaching, teaching, learning, and education there is a great diversity of approaches – and no shortage of strong opinions, judgments, and excuses. On a level applicable to all projects, this is explainable by difficulties of execution and communication, by differing missions and selfish choices, by resource constraints and conquering battles. But there are sources of disagreement and ineffectiveness that transcend ineptitude, selfishness, and malice and apply more specifically to mentorship. They are most apparent in parenting of small children so I’ll focus there.

Learning Areas

A major culprit is insufficient appreciation that learning experiences influence all of the following domains:

• Skills specific to the subject being engaged
• Meta-capabilities like perseverance, concentration, logic, introspection, empathy
• Values and Habits like honesty, compassion, independence, fairness
• Facts that define truths of the subject and/or the world
• Models that explain functioning of the subject and/or the world
• Resources like status, connections, credentials, memories
• Relationships with individuals, groups, subject, society, self

Mentoring approaches are optimizations that rank not only specific attainments, but these general areas. When this happens without sufficient awareness, interdependencies threaten unpleasant side-effects. When it isn’t made explicit enough, outsiders are tempted to presume ignorance and defend worthiness of sacrificed components. And even when deliberate and explicit, chosen trade-offs are naturally challenged by proponents of alternate rankings.

Over the short term, tactics and strategies necessarily prioritize specific lessons and learning areas. But improvement engenders opportunities and dangers elsewhere. Progress exposes new weaknesses and alters the critical path. All areas require timely care lest they compromise the others. Over the long term, there are no permanently optimal rankings and no permanently effective tactics or strategies.

Still, at any specific time and under any specific constraints there is a single optimal arrangement. The closer one gets to it the greater the potential effectiveness of execution. Fortunately, while perfection is valuable, it is not required for success.

Optimization

The best arrangement maximizes opportunities for progress across learning areas and minimizes collateral damage to areas not directly pursued.

Every such arrangement is temporary and sufficiently effective alternatives usually exist. It is thus more important to avoid the terrible than to find the perfect. But it is essential to recognize when fit turns poor.

Even the riskiest, most objectionable approaches can work well when coupled with acute awareness and skillful adjustments. Conversely, even the safest, most reasonable approaches assure mediocrity – or failure – when used for too long.

Uncertainty and change encourage optimizing for better results over many optimizations. They reward readiness to pursue goals most compatible with circumstances of the moment. They reward maximizing optionality so that more goals can be pursued under friendly circumstances.

How does one judge compatibility of an approach? To paraphrase Tolstoy’s quip about families, all compatible approaches are alike while all incompatible ones are incompatible in their own way. When areas are in balance and the approach well-suited to circumstances and needs, execution feels smooth. There are still challenges, objections, misunderstandings. There are still distractions, disagreements, fights, and bad days. But shared goals and good faith remains. Flow state is occasionally reached. The challenges are over the details and are usually approached cooperatively.

When fit is poor, disagreements disconnect from proximate causes. Fights feel rooted in deep and unstated distrust, disrespect, or disconnect. When left unaddressed, every task can begin to feel like a chore, every request can engender objections, every interaction can end in a fight. The real problem remains untouched after fighting ceases. ¹The movie Good Will Hunting offers examples such interactions. It is important to recognize first signs of disconnect, identify possible deeper causes, and adjust the approach to address them before they get buried too deeply. They are on the critical path now.

Coherence

The key to compatibility is in the foundations. While the optimal arrangement may be in constant flux, there is a deeper, more stable ranking that grounds long-term goals and motivates specific pursuits.

Is the ultimate goal domain excellence? A set of character traits? Belonging and happiness? Wealth and recognition? Understanding and usefulness? There are interconnections between goals like these that exert powerful, long-term influence on each other. But there are also conflicts that demand prioritization.

Care must be taken that shorter-term strategies and tactics remain a means subservient to, and compatible with, the chosen long-term ends.²The relationship between the two rankings resembles one between practitioners and theoreticians. It is essential, but difficult, to gracefully let go of approaches that cease to be optimal; to maintain focus on ultimate, long-term priorities; to not let rewards, metrics, and sub-goals subsume the reason they were introduced. Unaddressed incoherence seeds disconnect.

But the deeper, foundational ranking is its own source of conflict. It is intertwined with communal norms and thus influenced by them. It is sensitive to motivations and time horizons and thus varied across students, parents, coaches, teachers, and institutions. While it is fairly stable in competent mentors, it evolves in mentees. And growth in competence itself affects this evolution.

Despite the challenges, maintaining coherence and connection are the essence of mentorship.

Action Hierarchy Building

While the most visible output of mentoring is seemingly isolated skills, much of the task is development of comprehensive action hierarchies. This seeds misunderstanding and disagreement. Rapid skill progress can be achieved by sacrificing the foundations. Conversely, fitting foundations provide the meta-skills, motivation, and trust that can make higher-level progress appear easy or lucky: a product of nature more than nurture, of circumstances more than agency.

The opportunity to develop coherent hierarchies is greatest in early childhood when mentors have maximum opportunity to control the experience, establish trust, and build from first principles. Their later efforts will be increasingly consumed and constrained by management of perverse incentives, unwelcome temptations, and contradictory values of an unsympathetic environment.

The impact of effort is also more apparent in early childhood. While the concepts apply to mentorship at all stages, the necessity of understanding and accommodating established hierarchies under difficult conditions exerts increasing influence over time.

Granted, it is difficult to allocate results between biology, mentorship, and external environment with confidence – even in early childhood. Still, it may help to consider whether objections would necessarily apply to a six-year-old raised in an environment optimized for eliminating them.

This doesn’t mean that children are Play-Doh that invariably molds to desires of a sufficiently skilled sculptor. It is improbable that a child can be turned into whatever one wishes: biological factors and imperfections of execution are inevitable. And, in any case, the environment can never be sculpted perfectly enough to allow it, especially since most meaningful goals and impactful pursuits require eventual interaction with the outside world.

But mentors can encourage some natural tendencies over others, sculpt the early environment sufficiently to affect order or direction of development, model behavior, choose methods that better develop early potential, and so on. The important achievement is less in excellence of predetermined high-level skills and more in development of crucial values and meta-capabilities, establishment of trust and motivation, construction of a sound early version of the action hierarchy. These foundations increase the chances of mentor’s continued influence, of excellence in a variety of skills, and of steadfastness in the face of contradictions and temptations of unsanitized reality.

The task demands clarity, consistency, and empathy. It is fundamentally applied philosophy and psychology above anything else. Philosophy to develop a realistic set of desirable principles and to resolve everyday questions and challenges in ways that fit coherently and resiliently into the hierarchy being constructed. Psychology to model the pupil accurately enough to understand their concerns and intervene effectively yet subtly enough to maintain motivation, cooperation, and trust.

When the process involves a few people in a fairly stable environment, it offers a testbed for coherence of philosophical and psychological principles. It is an experiment, but one that remains closer to exploration and invention than to validation. It progresses one mistake at a time and thus demands flexibility, ingenuity, and adventurousness. But it also demands conscientiousness to keep adjustments consistent with the long-term vision.

An approach coherent enough to produce noteworthy results doesn’t lend itself to explanation or scalability. As crucial, difficult, and rare as effective application of tactics and strategies is, the most precious skills of mentoring are in the art of modeling the pupil; identifying weaknesses and sources of motivation; finding ways to foster motivation while directing some of it towards critical areas; all while nurturing trust and maintaining buy-in. Think Young Lady’s Illustrated Primer. Or Scott Alexander’s lament that Laszo Polgar’s trick to raising geniuses is “monomaniacal focus, a lot of free time, and hard-to-define talent.” At high-enough levels it is indistinguishable from magic.

The pedagogical approach must itself fit with the action hierarchy being constructed. At its root is a fundamental philosophical choice: to rely on dominance, negotiation, or kindness.

Mentorship Philosophy

Diana Baumrind’s typology gives us three common, defensible parenting approaches: authoritarian, authoritative, and permissive. For completeness, Maccoby and Martin added a difficult to defend default: neglectful. I find this categorization applicable to mentorship more generally.

Each defensible approach can be viewed as a sophisticated action hierarchy grounded in different philosophical choices and psychological theories. Authoritarian parenting embraces the philosophy of hierarchy; authoritative of agreement building; and permissive of emotional attachment.

In actuality, each approach requires all three: capable and competent authority, reasonable and defensible structure, goodwill and connection.

Authoritarian and permissive approaches rely heavily on cultural norms to backstop components they take for granted and prevent degeneration into pure tyranny – of either the adult or the child. While they can work well in compatible cultures, they are fragile in incompatible ones. The authoritative approach is more self-contained and robust, but is correspondingly more difficult to implement. While it thrives in compatible cultures, it has a greater chance of surviving incompatible ones – or at least degrading more gracefully.

It is noteworthy that these components resemble the Three Pillars of a Stable State that Fukuyama develops in Origins of Political Order. Authoritarian approaches emphasize the equivalent of a Strong and Modern State; authoritative ones of the Rule of Law; permissive ones of Accountability.

The connection offers us independent examples of interactions between these components. They are historically full of competitiveness, misunderstanding, and hostility. Yet, according to Fukuyama, it is the approximately equal balancing of these conflicting forces that is the secret ingredient for a modern liberal democracy.

The authoritative approach strives for the greatest balance and embraces values, preferences, and trade-offs closest to those of a modern liberal democracy. But attempts at balance increase exposure to external misunderstanding and hostility in parenting as they do across fields.

Like a well-functioning state, an authoritative parent must sometimes be forceful or harsh. Force and undesirable natural consequences ultimately ground all rules and agreements. To not follow through on a threat, or to readily soften the natural consequences, is to undermine the Rule of Law.

This doesn’t mean that fear of undesirable consequences must be the proximate driver of good behavior – one hopes that the desire to do the right thing and be a good person can do the heavy lifting. But the long chain of “what happens if I/they don’t do the right thing” questions that justifies good behavior ultimately ends in undesirable consequences.

Like a well-functioning state, an authoritative parent must sometimes be trusting or forgiving. Connection requires faith. Understanding depends on goodwill. Learning needs mistakes. To disallow freedom, ignore context, or punish indiscriminately is to undermine Accountability.

What makes exercise of force different from tyranny, natural consequences different from apathy, and lenience different from permissiveness is their perceived fairness, their consistent integration with the totality of circumstances, their deliberateness.

The parent and child act within the context of previous agreements and interactions. An effective authoritative parent vigilantly lives up to their end of the bargain – and holds the child accountable to theirs. They deliberately structure the environment to nudge the child to the edge of their abilities while offering choice and freedom. They are acutely aware of the recent history of demands, victories, and failures; of the current strength of the relationship; of their child’s psychological state – and adjust interactions and expectations accordingly towards excellence, fairness, or sympathy.

External Misunderstanding and Influence

The underlying sophistication only makes outsider misunderstanding more likely. This isn’t mainly due to inherent difficulties of understanding complex action hierarchies because most interactions with the outside world aren’t prompted by philosophical curiosity. They are prompted by interpretations of isolated events and these interpretations predictably assume simple extremes.

Although deliberate planning, respectful dialogue, and careful implementation dominate in time and effort, they don’t attract outside attention. It takes a tuned and interested observer to discern attentive listening, gentle suggestions, and subtle redirects of constructive engagement – and to perceive deliberate hierarchy building within such minor interactions. Like in a well-functioning state, components of long-term effectiveness go unnoticed.

But explosive events bring attention. Harsh actions brand parents as authoritarian. Lenient actions brand parents as permissive. Indeed, the actions themselves aren’t necessarily different from what authoritarian or permissive parents would do; the difference lies in their justification and context, in the balancing of goals.

It is difficult to observe an event and not presume single-minded optimization by participants. Indeed, there is little wrong with making this presumption if one then proceeds to verify it by exploring motivations and understanding underlying systems. Or if one simply resists the urge to be a busybody.

While presumption of single-minded optimization misdiagnoses, it gives enough credit for competence to default to cordial questioning or non-interference. Defending choices – especially misdiagnosed ones – in the middle of a charged, nuanced, time-sensitive situation is a frustrating extra burden, but a manageable one. Righteous interference is more debilitating.

It comes from assuming neglect, incompetence, or extreme misapplication. Misapplied authoritarian parenting is out of control: it punishes to hurt the child, to provide an outlet for the parent, or because it doesn’t know what else to do. Misapplied permissive parenting is apathetic: it provides freedom to get the child out of the way, to allow the parent to express themselves, or because it doesn’t know what else to do. A well-meaning observer may want to help a seemingly out-of-their-depth parent or may feel obligated to help a seemingly endangered child.

Overwhelmed parents and abused children are common enough to encourage action. Unsurprisingly, presumptions of single-minded optimization and incompetence tempt opposing single-minded optimizations as solutions. If these get codified into laws, the obligation to interfere increases which pressures the definition of abuse and the scope of laws to expand further in a positive feedback cycle.³One example highlighted in The Coddling of the American Mind is the path from abduction of Etan Patz to creation of the National Center for Missing and Exploited Children to the World’s Worst Mom incident to the Free Range Parenting counter-movement. Another example (not in the book) is expansion of mandatory abuse reporting and encouragement of reporting by the public.

Complex hierarchies and competent, nuanced approaches end up as collateral damage. Authoritative parenting is not possible without strictness, but visible strictness risks being interpreted as authoritarian abuse which may engender difficult legal or social interactions. Authoritative parenting is not possible without trust, but visible freedom-granting risks being interpreted as permissive neglect which may also engender difficult legal or social interactions.

Nuanced balancing of conflicting optimizations necessary for complex, long-term goals like independence, resilience, and excellence has a hard time existing under zealous scrutiny. This isn’t specific to parenting: Seeing Like a State offers plentiful examples of well-intentioned, single-minded optimizations gone wrong.

Distrust of the Deliberate

While some confuse complex interactions for errors or simplistic optimization – or complain about optimization for the wrong thing – others object to optimization, coherence, or deliberateness as such.

One concern is with stifling child’s innate motivations, suppressing discovery and development of their natural preferences, damaging their autonomy, sabotaging their drive to self-actualize. But these are consequences of taking away the child’s power to express, experiment, and choose which are only indirectly connected to mentor’s optimization and deliberateness. Effective mentors can follow the child’s lead, resist outright prohibitions, and leave final decisions to the child.

With all the power in child’s hands, how do these mentors exert deliberate control? As Rousseau wrote in Emile, or On Education:

“Let him always believe he is the master, and let it always be you who are. There is no subjection so perfect as that which keeps the appearance of freedom…Doubtless he ought to do only what he wants; but he ought to want only what you want him to do.”

The mentor restricts options by controlling the environment, developing the child’s action hierarchy, and taking advantage of path dependence. The child feels respected and in control which aids engagement, forthrightness, growth, and autonomy.

Some are concerned that such attempts to sanitize the environment take away true freedom. Others find the manipulation ethically unsavory and believe that it will backfire once exposed.

But the key to implementation is the child truly having freedom and their choices truly being respected. This respect cannot appear contingent on mentor’s optimizations as rejected choices and curtailed freedoms will indeed seed resentment. When the child gleams options the mentor did not want them to see and makes choices the mentor did not want them to make, they must be allowed to proceed. The mentor can seek concessions, find positive lessons to teach in context, and seize opportunities to bring the child back when choices backfire or novelty wears off – but they must not take away agency. They must be willing to risk their entire enterprise rather than sacrifice choice and trust.

As this is dangerous and difficult, most practitioners place great emphasis on control over the environment. Nevertheless, they only want this control if it can be had without sacrificing the appearance of agency. Only once does Rousseau appeal to his authority or relationship with the pupil – and that is when Emile is an adult.⁴It may be impossible to appeal to authority or sympathy as little as Rousseau portrays, especially in modernity. Emile is a fictional character and Rousseau gave all five of his biological infants to an orphanage; the youngest was surrendered ten years before publication of Emile. The magnitude of the claim nevertheless emphasizes the importance of agency to the approach.

Still, the child’s choices are filtered and managed which may inadvertently optimize for local, outdated, or ill-fitting maxima. And manipulativeness feels uncomfortable ethically.

But what alternatives are there?

No Better Alternatives

Learning through a multitude of truly unsanitized experiences isn’t actually a possibility. To play outside is to be manipulated by norms of the local peer group – partially filtered by randomness of first encounters. To consume content is to be manipulated by creators and marketers – partially filtered by randomness of timing and external influences. To explore the internet is to be manipulated by algorithms of search engines and advertising networks – partially filtered by chance. To engage with any environment – whether toys, books, schools, playgrounds, or forests – is to be manipulated by what the environment exposes and espouses – and to be directed by initial conditions, first encounters, and path dependence.

Nor can parenting, mentorship, or even close interaction be free of limits, preferences, and pressures. To eschew limits is to abandon all expectations including those of politeness, fairness, competence, safety, and empathy. To avoid expression of preferences is to withhold genuine interaction, empower louder voices, and model apathy. To choose options on the fly and make intuitive corrections is to act on subconscious preferences driven by optimizations and manipulations of one’s psyche and influenced by the randomness of one’s momentary emotional state.

The child can never see all possible options, escape the influence of path dependence, or make choices free of external pressures. And the mentor can never withhold all preferences, act without an agenda, or exercise no authority. The alternative to mentor’s manipulative curations are manipulative curations of semi-random actors. The alternative to deliberate optimizations are subconscious or unnoticed ones. It is therefore not enough to present disadvantages of manipulation, optimization, or sanitation: one has to argue for advantages of randomness and incoherence.

Randomness and Incoherence

There are indeed advantages to them. A person raised in a world that is too consistent risks having insufficient flexibility or realism to operate in imperfect or adverse circumstances. A person raised with choices that are too sanitized risks being insufficiently comfortable with unstructured exploration, insufficiently capable of unguided decisions, and insufficiently confident in their sense of self.

But nothing stops these valuable meta-capabilities from being included among deliberate goals of mentorship. Nothing stops the mentor from allowing the incoherence in, even as they try to manage its nature, timing, and context. Nothing dictates their handling of inevitable surprises.

Conversely, nothing stops free-flowing encounters with the real world from pushing their own preferred consistency. Nothing stops the child from adopting an external action hierarchy with vigor their mentor wishes for their own or with dogmatism feared by detractors. And nothing stops the child from assembling randomness into coherence of an imagined kind.

Assumptions of order, consistency, and coherence are built deeply into the process of learning. We rely on them to generalize from our experiences and make predictions. It is difficult to break these presumptions without breaking learning and purpose. A person raised in a world too inconsistent risks insufficient sense of agency, risks insufficient drive to explore or choose, risks never developing a sophisticated sense of self at all.

Some believe that such ego-humbling turmoil seeds truer perspectives, that children will come out from it stronger, that protection is delusional or counterproductive. But if this is so, how are mentor’s attempts at consistency and control not just another force? If it is healthy to be assaulted by an untold number of unspecified forces – many of which are optimized, controlling, and manipulative – then surely mentor’s pressure cannot be so disabling. And if protection from it is beneficial, then how could protection from those untold others not be?

The Child’s Point of View

The essential thing is that the effect on the child does not follow from external intentions, but from the child’s interpretations. The child will seek out meaning, will find pattern and purpose, will learn lessons and build systems out of their experiences. They will do this whether the experiences are crafted deliberately, built on the fly, or occur randomly; whether they are a product of selflessness, carelessness, or deceit; whether they are attempts to teach, humor, distract, or abuse. The child will build an action hierarchy from their interpretation of reality, optimize themselves around it, and object to inconsistency and unfairness they perceive.

Non-deliberateness isn’t automatically more natural, child-directed, invigorating, free, or stable. It merely cedes influence. Nor does it automatically improve trust, connection, autonomy, or realism. The child will take for granted whatever is given and reach for more.

And deliberateness isn’t automatically authoritarian, lecture filled, or committed to structured lessons.⁵Emile, or on Education, though verbose and dated in parts, offers an extensive philosophical treatment of an alternative. This chart provides the highlights. Creativity, curiosity, and flexibility can be encouraged deliberately – and more effectively. Nor does it automatically deny the influence of biology, culture, or peer group. These constraints can affect the timing, focus, and style of optimization without affecting its value.⁶For example, The Nurture Assumption’s cogent argument that environmental influence is mostly that of peers, not parents can be interpreted as evidence of inherent limitations or incorrect approaches. Hold On to Your Kids, for instance, flips many common parenting tactics and sub-goals to counteract the powerful pull of peers.

Deliberateness isn’t without costs and risks, but it does not have to clear the bar of perfection: it only has to beat randomness. To argue against deliberate optimization one has to, in effect, argue against agency. They need to argue that subconscious optimizations and biases are superior to conscious ones, that uncritical acceptance of external optimizations and chance is superior to attempts to do better.

Non-Deliberateness

In at least two major respects they are significantly inferior – even if we ignore inefficiency. The haphazardly built hierarchy will be more magical and difficult to understand with all the miscommunication and difficulty that this brings. And ignorance of one’s subconscious optimization for natural preferences risks a more violent revulsion to their violation. Together, these tempt authoritarian restrictions, permissive guilt-tripping, and neglectful withdrawal – regardless of the initial intentions. By not attempting to manage the environment, the action hierarchy, and self-understanding to limit these risks, one makes them more likely and more dangerous.

On the other hand, even a mediocre plan executed reasonably is superior to an excellent plan executed erratically. Underneath intuitions and norms is some amount of coherence. And intuitions may have a greater genetic overlap with one’s biological child than a bolted-on system. Intuitive actions have the benefit of being easier to execute consistently.

The importance of execution may similarly justify trade-offs between coherence of mentoring and mentor’s time, effort, attention, sanity, and other resources. When skills or resources are insufficient, lowering standards may be the best one can do to maintain execution effectiveness.⁷For example, when caretakers of infants approach their wits’ end, common advice is to put the child in a safe place and walk away even if you would normally consider this terrible. The idea can be extended to ineffectiveness that does not concern safety. The well-intentioned mentor does need to take care that such enticing justifications don’t slip into excuses and self-fulfilling prophecies.

High Modernism

Some argue that deliberate system-building is so complex that even the most capable and well-equipped are liable to overextend and fail in ways that overwhelm the benefits.⁸Off the Charts showcases the complexity and unpredictability of childrearing with stories of prodigies. Perhaps deliberateness does not have to be authoritarian, but mentors nevertheless succumb to the urge to prohibit choices that endanger success or fail to rein in their growing hubris when things go well.

This is concern with high modernism that Seeing Like a State covers so well: with reductionist over-optimization for what is visible, with disregard for complex trade-offs embedded in working systems, with downplay of higher-order effects, with dismissal of evolutionary change. This concern is a skeptical inverse of naive optimism, one that risks replacing energizing ignorance of trade-offs and hard-to-see effects with debilitating preoccupation with them.

As we’ve seen, the alternative to mentor’s deliberate optimization isn’t learning through a multitude of unsanitized experiences. It is a semi-random encounter with optimizations of others. And, in modern life, many of these optimizations are bona fide high modernism.

Furthermore, the risks of high modernism, over-optimization, and hubris being high does not imply that risks of status quo, incompetence, and mediocrity are low. There being wisdom in imperfection does not make striving for imperfection wise.

Nevertheless, there are differences between these types of risks and some objections rest on them.

Striving, Social Fit, and Risk Aversion

Much deliberate optimization strives for absolute limits: individual potential, skill excellence, truth, morality, efficiency, contribution, and so on. It argues that with a proper approach and sufficient effort we can get much closer to the limits than many believe is possible. It deems that making this progress is much better than not; that it is worthy of effort, sacrifice, and risk; that the key task is finding approaches that work. It bemoans potential lost to conventions.

But many strive for individual well-being. They argue that social fit and contentment are more important for happiness than endless exertion towards limits. They are less concerned with undeveloped potential, mediocrity, biased knowledge, dubious ethics, inefficiency, freeloading, and so on than with abuse and unhappiness. What they bemoan is maladjustment.⁹Objections to tiger parenting galvanized by the Battle Hymn of the Tiger Mother frequently have this flavor.

Belonging is itself an optimization that can be pursued with varying degrees of deliberateness. But it doesn’t feel like one since adherence to norms aids integration so naturally. What manifests as an objection to deliberateness, or as concern with specific risks, is often a discomfort with deviation from conventions.

Adherence to norms isn’t without risks and may well not optimize one’s chances of fitting in. Local conventions are unlikely to match ones at locations where mobile lifestyles may take the child; rapid change assures obsolescence of all norms; the best long-term chance of rapport may well require pushing for limits. And while belonging aids happiness, it does not assure it: successful striving and honest self-assessment also aid happiness and arguably offer a more sustainable path towards it.

Nor does striving inherently preclude social fit: strivers desire recognition and comradery like everyone else. That striving does not naturally lead to acceptance and happiness is at least as much a consequence of non-achievers optimizing for their own belonging as it is of achievers sacrificing fit. It is the demotion of excellence, integrity, and truth that makes misfits out of strivers, not striving itself. Fortunately, the cohesive social world is small and the physical world offers a great diversity of them. To adapt or rebel aren’t the only options: one can withdraw to seek small worlds with better fit.

Nevertheless, adherence may be better on average than any deliberate deviation from convention. Even if it assures mediocre outcomes, it reduces the possibility of horrible ones. And, in any case, it leaves one with plenty of sympathetic company. This company is a source of comfort and power. If you follow norms you will never be alone, will always have someone to blame, and will have the numbers to turn this blame into concessions. It isn’t a noble path, but it is a safer one.

Indeed, the aversion to risk and responsibility may underlie most objections. To deliberately optimize for positive goals is to define what matters and how to achieve it; to accept responsibility for the quality of execution and for the trade-offs inevitable to any ranking; and to risk standing alone. Conversely, to make the seemingly prudent choice and optimize for not doing harm is to meekly accept the default level of harm; to absolve oneself of responsibility even if the absolute harm level is greater.

Regret is a Terrible Signal

A noteworthy source of risk aversion is real or imagined regret. Negative experiences, or fears of them, ascribe a particularly high cost to perceived causes and engender an especially strong desire to avoid them. But regret only tells you what not to do. It tempts extreme, reductionist over-optimization that perversely inverts the Anna Karenina Principle: x made my family unhappy therefore avoiding x is required for happiness.

Regret is a terrible signal because avoidance focus invites reciprocal failures. Success demands a reasonably holistic, positive plan and such plans require committing to trade-offs, accepting risks, and balancing possible regrets.

A lighter version of the same false dilemma stems from positive experiences: I was happy because I did x therefore I would not be happy if I did y instead. But one’s happiness is a product of fit between identity and experience and earlier experiences influence identity formation. Doing y instead of x affects happiness standards.

Every Experience Educates

Which brings us back to where this post began: the variety of areas affected by learning experiences. One is never merely learning a skill, feeling joy, understanding a fact, bonding, playing, building confidence, gaining accolades, making sense of the world, developing empathy…

Every experience affects all learning areas. Every experience shapes the action hierarchy. Every experience is a learning experience whether it is initiated by the child, mentor, or happenstance; whether it is free-flowing, carefully planned, or chaotic; whether it is welcomed, resisted, or abhorred.  

To see mentorship too reductively is to sacrifice much opportunity for short-term optimization, long-term effectiveness, and understanding of successful mentors. It is the multitude of goals and influences that justifies the confidence to choose one most compatible with the immediate circumstances. It is their unequal importance to long-term goals that justifies the patience to prioritize fundamental values, models, and meta-capabilities over glamorous, short-term accomplishments. It is the quality of such decisions that underpins success.

But to see the endeavor as too amorphous to be understood, influenced, or optimized is to give up similar ground. It is the connection to influence that justifies the increased effort and sacrifice of quality optimization and prioritization. It is their outsized importance to long-term influence that makes trust, connection, and values so critical.

One can find influence less important, or consider it too dangerous or difficult to wield, or believe that external factors are too strong to be worth trying to overcome. Which goals to pursue, how much effort to expend, and how much opportunity to leave to chance is for each person to decide.

Any approach can be separately evaluated on its desirability, plausibility, accessibility, and cost. But whether mentors choose active involvement or not, whether they optimize deliberately or not, whether they execute competently or not – every experience educates, for better or worse.