Swartzism

Version 0.0.1: A Request for Comments

Wed, 14 Oct 2026 00:00:00 +0000

Today the doctrine ships — not as a finished thing, but as an opening position. Version 0.0.1 of Swartzism is published as a Request for Comments: five axioms, an ethics of scale, one boundary, five works, and a process for changing all of it except the last.

Read the Manifesto. Then read the Principles — the analytical machine underneath: every claim about the record carries a source, and one fabricated citation has already been caught, corrected, and logged. Then read the Governance — how the text changes, what never changes, and why.

Three commitments come with the version number:

The boundary is constitutional. Liberate the world’s knowledge; defend each person’s life. Everything else in the text is revisable by argument. The boundary is not revisable by coalition, cost-benefit, or mood.
The text must survive its own attacks. A version that has not survived a published attempt to destroy it is not stable, whatever its number. The review log is the doctrine eating its own cooking — and Pass 2 is now public business.
No one owns it. The doctrine is apersonal: arguments are judged, not authors. It publishes under its own name, takes help, not direction, and carries no signatures.

The profession test governs its texts: a doctrine that taxes its professors shrinks to a church of martyrs. This one intends to grow — which is why 0.0.1 is an invitation, not a proclamation.

Attack it where it is weakest. That is what the comment process is for.

The PII Question: How Swartzism Treats Data About People

Mon, 03 Aug 2026 12:30:00 +0000

The objection to any information-liberation doctrine writes itself: if you’re for freeing data, you must be fine with leaking people’s data too. Swartzism answers with a boundary that does not move:

Liberate the world’s knowledge. Defend each person’s life.

Data about the world and data about a person are different kinds of things. They are not both “data.” The doctrine is deliberately asymmetric about them.

The two-column test

Every data project, leak, or archive gets sorted into one of two columns:

Column	What it is	Swartzism’s rule
World-data	Dockets, gazettes, scientific literature, corporate registries, government records	Liberate. The public cannot govern what it cannot inspect.
Person-data	Names, biometrics, financial records, medical histories, locations, communications	Protect. The holder of this data holds power over the person it describes.

The first question about any contested dataset is not “should this be public?” — it is “who does this data describe?” The test is not “was this information obtained legally?” — the PACER download was a Terms-of-Service violation, and Swartzism still treats it as justified. The test is: whose data is it, and what does exposing it do to the person it describes?

Why the boundary is load-bearing

The doctrine has no chance without trust. Institutions will always argue that openness is a slippery slope to surveillance. Every time someone calling themselves a “liberator” dumps ordinary people’s data, they prove the institutions right. The boundary is not a policy choice within Swartzism; it is the load-bearing wall of the whole argument. It does not move.

The asymmetry is the point, not an inconsistency: the powerful’s secrets are ours to expose; the weak’s data is theirs to keep. A custodian who hoards knowledge is the enemy. A liberator who leaks lives is worse.

Leaks that fail the test

Consider a leaked customer database: millions of bank records, phone numbers, and addresses dumped to the open internet. The thief may call it “exposure.” Swartzism calls it what it is — the transfer of personal data from one holder to a thousand holders, each of whom can now impersonate, defraud, harass, or sell. ¹

A PII leak is not civil disobedience. Civil disobedience targets a structure of power; a PII dump targets the people the structure was already mistreating. It amplifies the harm rather than dismantling the cause. The victims of a bank or Aadhaar-style breach are not the institution — the institution has already been paid to care. The victims are the account holders whose data now circulates as currency in fraud markets.

When a leak of personal data surfaces, the doctrine’s instruction is clear:

Do not redistribute. Not one row. The data describes persons; propagation is the second harm.
Work for takedown and remediation. The affected people come first.
Strike the system that collected and exposed it. Investigate the custodian, the contract, the security failures, the regulator asleep at the switch. That is where the accountability work lives — in the system, not in the rows.

This is not timidity; it is the boundary working as designed. The research targets the leak as an event — who collected it, how it was secured, who failed, who profited. Never the people inside it.

The hard case: accountability data about the powerful

There is one genuinely hard case: official misconduct that is only provable through personal data. A minister’s abuse documented via records that include personal details.

The line has to be drawn per-case. A police officer’s substantiated misconduct finding is a public record about the exercise of public power; publishing it is world-data. ² The same officer’s home address, family, or financial details are person-data; publishing them is a target painted on a human being. The difference is the difference between the badge and the man. Swartzism holds the line at the badge.

Where release is the only path to accountability, the doctrine permits it only when all four conditions hold:

Imminence and gravity — the harm being exposed is ongoing or severe;
Minimality — the least personal data that proves the act;
Asymmetry — the subject holds power over the people affected;
Process — a documented proportionality decision by more than one person, reviewable later.

The purpose of these conditions is not to block release. It is to make the release defensible — because an accountability act that harms the innocent discredits the doctrine, not just the act.

The consequence

This is why Swartzism is not “piracy with a sad story attached,” and why it is not naive about power. It takes a side that confuses both the corporate enforcer and the careless “liberator”: knowledge of the world is a commons, and the privacy of persons is absolute — and the two commitments are one doctrine, not two.

That discipline is what lets the practice projects — the police-accountability archive, the hospital registry audit, the exam-tech investigation — do their work without becoming part of the problem they document.

The next stress test: what happens when the two columns collide inside a single system — an Aadhaar, a UPI, a national registry. That is where the doctrine meets Indian DPI, and where it will have to earn its keep.

https://en.wikipedia.org/wiki/Data_breach ↩
https://www.aaronswartzday.org/bad-apple-announcement/ ↩

What Is Swartzism?

Mon, 03 Aug 2026 12:25:00 +0000

The word does not exist in the record. Aaron Swartz never named a movement, never wrote a manifesto that founded an “-ism”, never claimed disciples. “Swartzism” is a construction — a name for the question: what does the record imply?

So the honest opening move is to define the term the way this blog will use it, and then say what it is not.

The raw material

Swartz left behind an unusually legible body of work: essays on Raw Thought written between 2002 and his death in 2013, the Guerrilla Open Access Manifesto (2008), the open-access advocacy, the PACER protest with Carl Malamud, the JSTOR mass download of 4.8 million articles that led to his prosecution and suicide, and the Semmelweis essay on how hard it is to see what is in front of you. (The Bad Apple police-accountability tool suite, launched in 2021 by the Aaron Swartz Day Police Surveillance Project, carries his name and ethos — a separate project, built in his memory.)

A doctrine built from this record has to be built from it: every claim traceable to a source, every failure admitted.

Three axioms, one boundary

The doctrine this blog will argue for rests on three claims and one boundary.

1. Knowledge about the world is a commons. Facts, laws, research, records — these are produced collectively and their enclosure is a harm. The manifesto was about scientific literature, but the logic extends: when a society’s own records (court filings, gazettes, registries) are priced and gated, the society is taxing itself on its own memory.

2. Information is power, and power must be accountable. The PACER action treated the free flow of judicial records as a form of accountability. The Bad Apple project treated the documentation of police misconduct as accountability. Information-as-power cuts both ways: it is why institutions hoard it, and why publics should pry it loose.

3. Privacy is a power too — held by the powerless. This is the boundary, and it is load-bearing. Data about persons is not a commons. The person described has a claim that the institution describing them does not. When a leak contains personal data, the doctrine’s instruction is not “liberate it” — it is protect it, and investigate the system that leaked it.

That last point is the one people get wrong about Swartzism, so it deserves its own post. Read it next.

What Swartzism is not

Not hagiography. Swartz was arrested in 2011 after a physical confrontation over a hidden camera at an MIT network closet — an escalation that did not serve his cause. A doctrine that requires its subject to be flawless is an idol, not an argument. The record includes the failures; the argument has to survive them.
Not “steal everything.” The boundary above is the entire point. Swartzism without the PII boundary is just piracy with a sad story attached.
Not a cult. No membership, no orthodoxy, no leader. An argument that can be stress-tested is the only kind worth having.

The name is a convenience. The argument is the thing.