← back to writing

Legibility Without Response

The new civic ritual.

essay / systems / answerability / legitimacy

A modern ritual goes like this:

You gather proof that you are real.

You scan documents that contain your face, your history, your relationships, your income, your compliance. You upload them into a portal with an interface designed to make you feel like the system is listening.

You pay the fee. You get the automated confirmation that your request has been received.

Then you wait.

Weeks become months. Months become something softer and more dangerous: indefinite.

If you are lucky, the system gives you a status word. Pending. Under review. In process. A noun pretending to be an answer.

You cannot speak to anyone who can actually decide. You cannot learn what the system believes about you. You cannot discover what would move your file forward, because nothing is presented as a file anyone holds.

Meanwhile, you are not invisible.

The system sees you. It logs your submissions, your travel, your addresses, your taxes, your digital exhaust. It is perfectly capable of recording your life at high resolution.

It just cannot, or will not, answer you in human time.

I know this ritual intimately. I am a U.S. citizen. My son is a U.S. citizen. My wife is Iranian. We have done everything we were told to do. We’ve paid for the privilege of being examined, and then we’ve lived inside the silence.

For a while I treated that as a personal nightmare. Lately I’ve realized it is a template.

A smaller version plays out every day in the algorithmic public square:

Your post stops distributing.
Your reach collapses.
Your account gets “limited.”
Your appeal disappears into a form.
Your job application goes quiet.
Your loan rate shifts.
Your insurance premium jumps.
Your review is filtered.
Your bank flags a transfer.
Your platform tells you the decision was made for “quality” or “risk” or “safety” or “relevance.”

Those aren’t explanations. They’re masks.

That experience used to belong mostly to the edges, the highest-stakes bureaucracies, the places where power was concentrated.

Now it’s everywhere.

And the more we modernize, the more normal it becomes.

The variable we forgot to measure

We measure a lot in modern societies.

We count growth, engagement, inflation, unemployment, productivity, test scores, creditworthiness, risk scores, compliance rates, platform impressions, conversions.

In the last year alone, you could fill a bookshelf with essays arguing that our official metrics are broken. Bad benchmarks. Wrong rulers. Narratives that don’t match lived constraint.

Even if you dispute the exact numbers in those pieces, the underlying point is hard to miss: when a society measures the wrong things, it starts optimizing against its own people.

But I think we’ve missed an even more basic variable, because it doesn’t show up on dashboards. We treat it like a mood instead of infrastructure.

Agency.

By agency, I do not mean control. Control is rare. Humans have never had much of it.

I mean something more modest and more structural:

Agency is the ability to take a reasonable action and receive a timely, legible response you can learn from.

A response can be “yes.” It can be “no.” It can even be “not yet,” as long as “not yet” comes with a clock and a reason.

Humans are calibrated for learnable feedback loops. When feedback is legible, we can endure hardship. We can adjust. We can plan. We can make tradeoffs that feel like tradeoffs rather than traps.

When feedback becomes illegible, effort starts to feel like theater. Compliance starts to feel like humiliation. Persistence starts to feel like self-erasure.

This is why so many arguments about “the vibes” miss the mechanism.

The problem is not only that people feel squeezed.

It’s that people increasingly feel unanswered.

The answerability gap

Here is the modern contradiction I keep running into:

Legibility to power has increased faster than responsiveness from it.

We are easier to watch than ever. We are harder to answer than ever.

This is not because everyone inside an institution is cruel. It is because scale changes what is cheap and what is expensive.

At human scale, observation and response are coupled. If someone can see you, they can usually answer you. Even a refusal is still an interaction that locates you in the social world. It tells you who has authority, what they think, and what you can do next.

At mass scale, observation becomes cheap. Response becomes expensive.

A system can ingest your information, score it, route it, and store it indefinitely. A human decision that is accountable, contextual, and timely is far harder to produce. It introduces discretion, liability, inconsistency, appeal risk, slow labor, and precedent. It is costly. It is difficult to audit.

So systems evolve toward what they can afford:

That widening mismatch is the core phenomenon of the age. Call it the answerability gap: the growing distance between what systems can see and do, and what they can answer for in a way that meets a human need for actionable feedback.

A related idea is responsiveness debt. Like technical debt, it accumulates when systems ship capability faster than they ship recourse. The debt stays invisible until enough people hit the wall at once. Then it shows up as rage, distrust, superstition, withdrawal, and periodic eruptions.

You can feel this in everyday phrases that have become normal:

When enough people are living inside those sentences, you don’t have a customer service problem.

You have a legitimacy problem.

A systems scientist’s view of the social climate

Earth systems scientists make sense of complexity by tracking a few slow reservoirs and the feedback loops that move them.

Here’s a human version of that model.

Slow reservoirs

Modernization increases scale, complexity, and legibility almost automatically. It does not automatically increase answerability, cohesion, or legitimacy. In some regimes, it actively decreases them.

Fast weather variables

These don’t create the climate. They ignite what the climate has been accumulating.

The central loop

  1. Scale rises
  2. Complexity rises
  3. Institutions replace judgment with process to manage variance and liability
  4. Answerability lags
  5. Perceived agency declines
  6. People substitute high-feedback behaviors
  7. Conflict rises, trust falls
  8. Institutions respond with more risk control
  9. Legibility increases further, answerability remains scarce
  10. The loop tightens

This is what “modernization” looks like when you plot it as a coupled human-technical system.

Once you see the loop, a lot of modern behavior stops looking like a moral collapse and starts looking like adaptation.

Not noble adaptation. Not pretty adaptation. But predictable.

A hundred years of modernization in six regimes

This is not a moral narrative. It’s a systems narrative. Each regime has a dominant coordination technology and a characteristic set of feedback loops.

Mass industry and the administrative state

1920s to 1945

Coordination technologies: industrial production, mass bureaucracy, mass media, national mobilization.

Large institutions learn to coordinate millions, especially under crisis. Paperwork becomes a governance substrate. The desk becomes a locus of power.

Answerability still has human boundary layers. Clerks exist. Managers exist. Neighborhoods have memory. A lot of life is still negotiated in person, inside institutions that can see you and answer you in the same social space.

The postwar bargain and thick intermediaries

1945 to early 1970s

Coordination technologies: stable employment structures, expanding public investment, thick civic institutions.

Many people experience a relatively tight loop between effort and security. Intermediaries are strong: unions, churches, civic organizations, local media, extended family structures.

This world is not equitable. It is often exclusionary. But for many, it is psychologically stabilizing because the feedback loops are legible. If you hit a snag, there is a person, a counter, a union rep, a friend of a friend. Even when the system is large, it feels livable because there are human bridges.

Shocks and technocratic substitution

1970s

Energy crises, inflation, restructuring. The old bargain weakens. Institutions lean more heavily on technical management. Policy language displaces moral language. People are asked to trust models and managers they cannot see.

The loop between citizen and institution lengthens. The first legitimacy fractures appear. Decisions feel more distant, more abstract, more framed as necessity rather than choice.

Optimization governance and procedural life

1980s to 2008

Coordination technologies: metrics, standardization, global supply chains, financialization, risk shifting.

What can be measured becomes what matters. Variance becomes the enemy. Humans become inputs to be routed. Liability becomes a design constraint. Discretion becomes a risk.

Answerability declines as process grows. Rulebooks expand. Phone trees replace people. Exceptions become harder. Appeals become theatrical.

At the same time, more of life becomes ticketed: childcare, healthcare navigation, credentialing, basic participation in the economy. The cost of being a functioning adult rises, while the legibility of how to successfully do it declines.

Legitimacy fracture and the attention layer

2008 to 2019

A major credibility shock. People watch institutions bend rules to save themselves, then insist those same rules are immutable downward. Trust decays.

Meanwhile, smartphones and social platforms turn the comment layer into a primary arena of status, identity, and belonging. Feedback becomes faster, louder, and more moralized.

Even as communication becomes instant, explanation and recourse do not.

Visibility starts to substitute for efficacy. Moral intensity becomes a form of traction.

Pandemic acceleration and ambient scoring

2020s to now

Remote becomes default. Platforms become infrastructure. Enforcement becomes ambient. Scoring and surveillance expand. Rules change without published diffs. People live under algorithmic weather.

Answerability does not rise proportionally. In many domains, it gets worse.

You can submit everything, be seen everywhere, and still not get a clock, a reason, or a reachable human with discretion.

From rules to weather

A crucial shift beneath the last regime is the move from governance by explicit rules to governance by models.

Rules can be stupid and still be legible. They can be published. They can be argued with. You can build a mental model: “If I do X, Y happens.”

Model-mediated systems are different.

A model produces an output based on a high-dimensional input space, often inside a pipeline of multiple models and policies. It may be stable at the population level and still be unnarratable at the individual level in a way that meets a human need.

This is the foreshadowing we are already living through:

Even experts and builders often cannot answer the question a human is actually asking.

Not “what label did the system apply,” but:

In many ML-mediated systems, the honest answer is too complex, too uncertain, too liability-heavy, or too gameable to be offered. So the answer becomes silence, or a category word that functions as a polite refusal to be answerable.

This is how “explainability” becomes a decoy.

Humans don’t need a feature attribution chart. They need actionable, accountable, time-bound response.

When that response is absent, people become forensic. They reverse engineer. They build folk theories. They treat the environment as adversarial weather. They become superstitious because superstition is a rational response to a system that punishes without publishing law.

Internet culture as a feedback machine

The internet didn’t only accelerate communication. It changed what gets rewarded.

In an attention market, what matters is not truth, but traction.

The attention escalation loop is simple:

  1. Engagement rewards intensity
  2. Intensity drives more engagement
  3. Platforms optimize for engagement
  4. Creators optimize for platform reward
  5. Baseline arousal rises

This loop doesn’t require evil intent. It requires metrics and optimization.

Now add monetization.

A few weeks ago I was talking to artist friends who decided to embrace contributor payouts on X. The requirement was not “make good work.” It was “get five million views.”

So what strategies emerge?

This isn’t “bad people doing bad things.” It’s rational behavior inside the incentive landscape.

The disturbing part is the mirror it holds up.

If this is the town square of a civilization, and the optimal move is to be fast, provocative, and occasionally wrong on purpose, then we built a square that selects for heat and pays it in cash.

Then we act surprised that everything feels hot.

Human adaptations when agency collapses

When people lose the experience of legible cause and effect, they do not stop needing agency. They substitute.

From the outside, these substitutions look like cultural pathologies. From the inside, they often feel like survival strategies.

Moral certainty

If you can’t change outcomes, being “right” can become the last lever that still moves something inside you. Certainty provides relief from ambiguity. It gives coherence. It gives social protection in polarized environments.

This is why moral intensity rises as institutional efficacy falls. Certainty is traction.

Identity fusion

When procedural systems make people feel interchangeable, belonging becomes a rescue rope. Identity becomes the primary way people feel causally effective.

That becomes dangerous when threats to the group are experienced as threats to the self, because proportionality collapses.

Outrage and spectacle

Outrage produces immediate feedback. It reliably generates engagement, alignment, and visible reaction.

In a low-agency environment, high-feedback behaviors become addictive. Not because people are shallow, but because the nervous system is hungry for response.

When ordinary action yields silence, abnormal action begins to feel like the only lever left.

Withdrawal

Withdrawal is the inverse strategy. If nothing you do matters, disengagement becomes a way to stop bleeding attention into systems that do not answer.

This shows up as cynicism, burnout, depression, and retreat into private life.

Exit fantasies

When voice fails, exit becomes attractive. People dream of leaving the system rather than changing it. This can be literal (moving, opting out) or ideological (total rejection, “burn it down”).

Exit is a form of agency when participation feels like coerced consent.

None of this is a justification for harm. It is a description of incentives and adaptation.

Tail risks and the security ratchet

In climate science, extreme events are not just anomalies. They are signals that boundary conditions have changed: higher baseline temperature, lower damping capacity, tighter feedback loops, longer recovery times.

Human systems have their own tail risks. Most people will never go there. But the existence and frequency of rare catastrophic discharges tell you something about the climate.

One predictable pattern in low-answerability environments is that people start seeking feedback by force.

Not because violence is “inevitable.” It isn’t.

Because when ordinary channels feel dead, the temptation to manufacture a response grows. People look for levers that cannot be ignored. They look for actions that will not be met with “pending.”

When rare tail events occur, they trigger a security ratchet:

  1. Fear rises
  2. Surveillance and control expand
  3. Answerability declines further
  4. Legitimacy decays for some
  5. Grievance increases
  6. Future tail risk grows

A few rare events can reshape governance for everyone. That is why answerability is not academic. It is a stability question.

Early warning signals of a regime shift

Earth systems scientists look for signs that a system is approaching a tipping point: rising variance, rising autocorrelation, slower recovery after shocks.

Social equivalents might look like:

These are not proof of collapse. They are indicators of declining damping capacity.

What happens next if the loops continue

If we keep the same feedback loops, the future is not one clean dystopia. It’s a set of converging trends.

Scoreboard civilization

More life will be mediated by scores, rankings, and dynamic thresholds: credit, trust, risk, safety, quality, employability, visibility.

This increases legibility. It does not automatically increase answerability.

People live as performers for models they cannot see.

Shadow-patch governance

Rules change without published diffs, not only on platforms but in labor markets, insurance pricing, lending, eligibility, and enforcement.

Experts become meteorologists of systems they can observe statistically but cannot narrate at the level a person needs.

The public treats governance as weather.

Permanent probation

Continuous evaluation replaces discrete gatekeeping. Access degrades softly: throttled, delayed, “under review.”

That produces a background state of hypervigilance and learned helplessness.

Intensification of the attention market

If money and status continue to be paid for impressions, outrage remains a yield strategy. Piggybacking and provocation remain rational.

The public square keeps selecting for heat and calling it culture.

More exit, less shared space

As answerability declines, the cost of participation rises. People sort into exit strategies: geographic, digital, social, ideological.

Shared institutions thin. Parallel institutions grow. The commons becomes brittle.

More tail risks, tighter security ratchets

A hotter social climate does not guarantee catastrophic events. It increases their probability and worsens their consequences.

Each tail event becomes fuel for the security ratchet, widening the answerability gap further.

A legitimacy crisis that looks like exhaustion

The end state is not necessarily revolution. It may be chronic disengagement: lower trust, lower participation, lower belief that voice matters, higher acceptance of coercion or spectacle as the only levers that work.

A society can survive hardship. It can even survive inequality for a long time.

What it struggles to survive is a sustained mismatch between legibility and response, between being seen and being answered.

What people mean when they say “the system is broken”

Often, they are not making a technical claim. They are describing a lived contradiction:

Nothing I do seems to change anything, but everything I do is still recorded.

That is not freedom. It is exposure without agency.

What would it mean to rebuild answerability

I don’t think the solution is nostalgia. I don’t think we can shrink society back down to the size where everyone knows the clerk.

But we can make a more modest demand, and we can make it measurable.

If a system is going to govern people at scale, it should provide at least three things by design:

Clocks you can trust

Not vibes. Not “under review.” Actual time commitments, especially in high-stakes domains.

If the system cannot commit to a timeline, it should be forced to admit that it is exercising power without accountability.

Reasons that are actionable

Not “risk.” Not “quality.” Not “relevance.”

Reasons that tell a person what the system believes and what could be changed, in language that a normal human can use.

If a reason cannot be stated without revealing trade secrets, then the system should be constrained from making high-stakes decisions about human lives.

Recourse that can change outcomes

Not a form that disappears into a void.

A reachable path with authority to review and reverse decisions, with an escalation ladder and a record.

Recourse is not a nice-to-have. It is what prevents people from turning to substitutes for agency.

Patch notes for rule changes

If rules change, publish the diff.

We have normalized shadow patches in the systems that govern daily life. Then we wonder why people act like the world is rigged.

A society that wants legitimacy should not govern by surprise.

None of this fixes human conflict. None of it eliminates bad faith.

It simply rebuilds the most basic condition for sane participation: legible cause and effect.

The conclusion that keeps reappearing

Across decades, across crises, across domains, the same unresolved variable returns.

When ordinary action stops producing legible results, humans do not stop needing agency. They just look for it elsewhere.

So the question is not how to shame people into better behavior, or how to dismiss the mood as delusion.

The question is:

Where do humans turn for legible, proportional, non-destructive agency in a world of massive systems?
Where can effort reliably translate into movement, even if it is slow, partial, and imperfect?
Where can people feel causally effective without having to become extreme, performative, or cruel?

If we keep the same feedback loops, the answer will increasingly be: anywhere that provides immediate feedback. Identity. Outrage. Exit. Rituals of certainty. Personal optimization for invisible models. Parallel realities.

Not because people are worse than before.

Because the environment selects for those substitutes when answerability is scarce.

Humans do not break because they have too much freedom.
They break when their actions stop connecting to reality.