Perfection is achieved when there is nothing left to remove

Via negativa is the discipline of improving life by subtraction rather than addition. Instead of asking “What should I do?” you ask “What should I stop doing?” The shift sounds simple, but it changes how you see progress. It treats many problems, such as personal habits, professional systems, and social dynamics, as fragile. In fragile systems, extra parts often create extra failure modes. Adding a new tool, a new commitment, a new metric, or a new rule can feel like forward motion while quietly increasing complexity, dependency, and risk.

The mindset is not anti-growth or anti-ambition. It is a bias toward removing what is demonstrably harmful, unnecessary, or misleading before you try to “optimize” what remains. That makes it especially practical when you face uncertainty, incomplete knowledge, or the temptation to chase novelty. You can be wrong about what will help you, but you can often be right about what is hurting you.

Below is a decision framework for applying via negativa deliberately, so it becomes more than a slogan, plus a set of caveats to keep it from turning into cynicism, fear of change, or status-quo worship.

Step 1: Define the outcome and the boundary conditions

Start by naming the outcome you want and the constraints you must respect. “Improve my health” is vague; “increase energy and reduce afternoon crashes without harming sleep” is clearer. In work, “be more productive” becomes “ship higher-quality work with fewer late nights.” In social or civic contexts, “fix the system” becomes “reduce harm while preserving what already works.”

Then write down boundary conditions: non-negotiables such as ethics, safety, relationships, legal constraints, time, and budget. Via negativa works best when you know what must not be traded away.

Output of Step 1: a short statement: Outcome + constraints.

Step 2: Map the system as it actually exists (not as you wish it did)

Before removing anything, understand what is currently in the system:

• Inputs: information sources, people, tools, routines, incentives
• Processes: how decisions are made, how work flows, how feedback arrives
• Outputs: what gets produced, what gets neglected, what gets rewarded
• Failure modes: how things go wrong (errors, burnout, conflict, drift)

This is where Chesterton’s Fence matters: don’t remove a “fence” until you know why it was built. That fence might be a meeting, a policy, a checklist, or a habit. If you can’t explain its original purpose, your first job is to learn—not to delete.

Output of Step 2: a simple list of components and their presumed purpose.

Step 3: Run the “removal candidates” test (what should not be there?)

Now generate subtraction candidates. The goal is not elegance; it is risk reduction and clarity. Look for:

• Unforced commitments: obligations you accepted by default (recurring meetings, subscriptions, social expectations).
• Complexity without proportional value: tools, workflows, or steps that add friction but rarely change outcomes.
• Metrics that distort behavior (Goodhart’s Law): targets that cause people to game the number rather than improve the reality.
• Toxic inputs: information diets that inflame anxiety, outrage, envy, or distraction.
• Ambiguity generators: unclear ownership, unclear definitions, unclear “done.”

Phrase candidates in negative form: “Stop checking X before noon,” “Stop adding new projects without closing one,” “Stop measuring success by Y,” “Stop responding to messages within 2 minutes.”

Output of Step 3: a ranked list of removal candidates.

Step 4: Apply a decision rubric to each candidate

For each removal candidate, score it on four questions:

1. Evidence of harm: What is the best evidence this causes harm (time loss, stress, defects, misalignment)? Is the harm recurring?
2. Reversibility: If you remove it and regret it, can you restore it easily? (Reversible changes are ideal early experiments.)
3. Second-order effects: What does it protect against? What dependencies rely on it? What new risks appear if it’s gone?
4. Asymmetry: Do you gain more by removing it than you risk losing? Via negativa seeks positive asymmetry—small removal, large benefit.

This rubric keeps subtraction from becoming reckless minimalism. You are not trying to remove “things.” You are trying to remove downsides.

Step 5: Design a small experiment (subtract, don’t demolish)

When possible, remove by experiment:

• Timebox removal (e.g., 2 weeks without a meeting).
• Reduce frequency before deleting (weekly → biweekly → monthly).
• Add a “circuit breaker” (conditions under which you reintroduce the item).

This is where reductio ad absurdum can help. Ask: “If we fully commit to this removal, what absurd outcome would appear?” If the answer is “we would lose critical safety checks,” then your experiment must preserve safety while still reducing waste.

Output of Step 5: an experiment plan: what changes, for how long, what you will watch, and what would cause rollback.

Step 6: Measure using anti-Goodhart metrics (and narrative evidence)

If the system is prone to Goodhart’s Law, prefer measurement that is harder to game:

• Fewer but higher-quality signals (defect rate, rework, sleep, stress, customer churn).
• “Lagging” indicators combined with short qualitative notes.
• Predefined check-ins: “What felt easier? What broke? What surprised us?”

Via negativa often improves the invisible: clarity, recovery time, cognitive load. Track those with brief journaling so your brain doesn’t rewrite history.

Step 7: Lock in the subtraction (or restore the fence with understanding)

At the end of the experiment:

• If removal helped, codify it. Make it the default. Remove temptation to re-add it casually.
• If removal hurt, restore it, but do so with explicit purpose, boundaries, and ownership. You didn’t “fail”; you learned the fence’s function.

Caveats: where via negativa can go wrong

1) Subtraction is not automatically wisdom

Removing the wrong thing can destroy resilience. Some redundancies are safety features: backups, peer review, rest days, and guardrails. If a component reduces catastrophic risk, it deserves extra scrutiny before removal.

2) “If it ain’t broken” is not a complete strategy

Some systems fail slowly: health, relationships, trust, security, and culture often degrade before they “break.” Via negativa should remove silent drift (small harms) and prevent future fragility—not preserve stagnation.

3) Beware moral licensing: minimalism as virtue signaling

It is easy to confuse “having fewer things” with “being better.” The point is outcomes: fewer problems, fewer failure modes, fewer avoidable regrets. If subtraction becomes identity, it can block necessary investment.

4) Subtraction can hide avoidance

Sometimes “stop doing X” is a mask for fear: fear of commitment, fear of feedback, fear of shipping. If your removals always target exposure (hard conversations, accountability, deadlines), you may be pruning the very actions that create growth.

5) Subtraction should respect other people’s constraints

In teams, what looks like “waste” might be someone else’s safety net. Removing a process without considering downstream users creates hidden costs. Apply the framework collaboratively, and treat dissent as data.

A practical summary

Adopting via negativa as a mindset means treating subtraction as your first move, especially under uncertainty:

• Define the outcome and constraints.
• Understand why the current fences exist.
• Identify what is clearly harmful or misleading.
• Remove through reversible experiments.
• Measure carefully to avoid Goodhart traps.
• Keep what works, restore what you misunderstood, and repeat.

Done well, via negativa becomes a quiet form of power: you stop paying for problems you don’t need, and you create space for what actually matters—without pretending you can design perfection from scratch.

References

Be aware of:

Chesterton’s Fence “Don't ever take a fence down until you know the reason it was put up." - aka Intellectual Humility link

Goodhart’s Law “when a measure becomes a target, it ceases to be a good measure”

Reductio ad absurdum, Latin for "reduction to absurdity," is a form of reasoning used to refute a proposition by demonstrating that its denial results in an absurdity or logical contradiction.

Nasim Nicholas Taleb: You know what’s wrong with more certainty rather than you know what is right or the Anna Karenina principle “all happy families are alike, all unhappy families are unhappy in their unique way”

Future IQ Podcast

Explanation from The Mind Collection