The Generator Function: Price's Law, the Mathematics of Manufactured Inequality, and What the Square Root Hides

From bibliometric artifact to platform algorithm — a Philosopher-Builder critique of the most fashionable productivity law of our decade

#price's law#structural dynamics#power#cumulative advantage#platform economics#commons#philosopher-builder

Publication: Sepahsalar.org/research Section: Political Economy / Structural Dynamics / Post-Techno-Capitalism Status: Working paper, open for collaboration License: CC BY-SA 4.0 Author: Ardeshir


“Mathematics doesn’t give a fuck about fairness or ethics. Compounding systems create exponential distributions, that’s it.” — Kaguura Gichuru, The Mathematical Reason Most People Never “Make It”

“The law is a knob. The knob is owned. The setting is a policy decision, not a discovery about human nature.” — The argument of this paper

The most fashionable productivity law of the decade is being deployed across Substack, LinkedIn, the founder-podcast circuit, and the operator-class self-help economy as if it were a finding in physics. Price’s Law — the square root of contributors in any domain produces fifty percent of the output — has become the favored alibi of platform capitalism’s winners. It launders concentration into competence. It launders algorithmic curation into “compounding advantage.” It launders four centuries of institutional enclosure into “this is just how complex systems work.” And it instructs the rest of us, in tones of avuncular concern, to identify our personal √n and “exploit it ruthlessly” before the late game closes.

This paper argues that the popular framing of Price’s Law is not wrong about the math. The math is real. What it is wrong about is everything else: the history of the measurement, the architecture of the systems being measured, and the political function of presenting an engineered outcome as a natural law. We hold the framing up against the actual record of where Price’s data came from, who built the institutions that produced it, and who currently profits from its mystification. Then we do the Philosopher-Builder turn: if the distribution is downstream of design, then the design is the lever, and the lever is what we are building toward in the Univrs Cloud Commons and the Mycelial Economic Network.

The popular framing is no longer interesting. The diagnosis is now the weather. What we build instead of it is the only question that matters.


I. What the Substack Is Actually Selling

The Kaguura piece is a competent specimen of a genre. Spotify Wrapped as hook. Derek de Solla Price as authority. A short, clean statement of the formula. Then the move: from “this is the pattern” to “this is how the world works” to “and therefore your job is to find your √n and double down ruthlessly.” The closing sections — Content, Skills/Offer, Time Management — are operator-class self-help, and they are useful at that level. Publish volume, track everything, identify what works, kill the rest. Fine. This is craft advice.

The problem is what the piece quietly does between the math and the advice. It performs three slides, each invisible to a reader who isn’t watching for them.

The first slide is from empirical observation to natural law. Price observed citation patterns in 1963. The piece reaches for astrophysics — the square root of stars produces half the light — and uses the analogy to suggest that human productivity inequality is a property of complex systems in general, on par with the stellar initial mass function. This is not an argument. It is a category error performed at speed.

The second slide is from natural law to individual ethics. Once the inequality is naturalized, the appropriate response becomes individual rather than structural. You cannot redesign the system; you can only identify your √n inside it. The whole back half of the piece is structured around this move.

The third slide is from individual ethics to legitimation of existing winners. If √n is real and natural and the appropriate response is personal optimization, then the people currently at the top of the distribution are simply the people who did the optimization. Their position is earned. Yours is, by symmetric implication, also earned. The 3,300 artists who get half of Spotify’s streams are the talented; the 10,999,996,700 who don’t are the untalented or the unfocused. Mathematics doesn’t give a fuck about fairness or ethics.

This is the politics being smuggled in under the math. We are not against the math. We are against the smuggling.

II. What Price Was Actually Measuring

Derek J. de Solla Price wrote Little Science, Big Science in 1963. He was not discovering a universal property of human productivity. He was describing the citation patterns of a specific institutional configuration: the post-war Anglo-American scientific establishment at the absolute peak of its deliberate state-engineered concentration.

The setting is not background. It is the entire story.

The Vannevar Bush memorandum Science, The Endless Frontier (1945) restructured American science funding around the NSF model. The Department of Defense, the Atomic Energy Commission, NASA after 1958 — all of these poured enormous capital through a small number of elite universities. The Soviet Union built the structurally identical apparatus through its Academy of Sciences. The Cold War R&D regime was the largest intentional concentration of scientific labor in human history, and it was designed to concentrate. That was the point. You don’t win a strategic competition by distributing research dollars evenly across the population. You pick winners.

The data Price counted had just been invented to make that picking legible. Eugene Garfield founded the Science Citation Index in 1955. Its coverage was, by deliberate design, weighted toward English-language journals at well-capitalized institutions in the Anglo-American sphere. Garfield was building a tool to help the system see its own concentration. Price counted what the tool reported. He found concentration.

A man counts the output of a funnel and discovers, to his evident interest, that more water comes out the bottom than out the sides. This is, schematically, what Price’s Law is.

The popular framing presents this as the discovery of a natural law. It is more accurately the audit trail of a specific historical project — Cold War big science — that succeeded at exactly what it was designed to do. The √n is not the shape of human capability. It is the shape of the institutional architecture that was running when the measurement was taken.

III. The Bibliometric Instrument Has a Politics

Citation counting is the foundational empirical practice behind Price’s Law and its descendants — Lotka’s Law of scientific productivity, Bradford’s Law of journal scattering, the h-index, the impact factor industrial complex. All of these rest on the same instrument. The instrument is not neutral.

Consider what citation indices systematically under-count.

Work published in languages other than English. Work published outside the journal system — technical reports, ministry papers, samizdat, internal industrial research, vernacular knowledge, oral traditions. Work by scholars at institutions without subscription access to the journals others cite. Work in fields the indices were slow to cover: most of the humanities for decades, most of non-Western philosophy and science, most of indigenous knowledge systems, most of the practical knowledge held by women, peasants, artisans, and the colonized.

Consider what they systematically over-count.

Work by scholars at institutions whose libraries hold the journals others cite. Work in fields with high citation density per paper — high-energy physics, molecular biology.

The instrument does not observe productivity. It observes legibility to the instrument. And the instrument was built by, and for, the institutions whose productivity it validates. Price’s Law inherits every bias of the citation index that produced it. The √n is not the square root of talent. It is the square root of institutional access to the measurement apparatus.

IV. The Generator Function Is a Design Choice

So here is where we are. The distribution is real. The concentration is real. The compounding is real. What is not real is the claim that these are properties of human nature rather than properties of the systems humans built and continue to maintain.

The √n is not discovered. It is generated. Every platform, every funding body, every algorithmic feed, every citation index is a generator function — a set of rules that produces a distribution. Change the rules, change the distribution. The question was never “how do I find my √n?” The question is: who set the function, and what does it cost the rest of us?

The popular framing forecloses this question by design. If the law is natural, there is nothing to set and no one to hold accountable. You simply optimize. The founders optimize. The platforms optimize. And the concentration deepens, wearing the face of mathematics.

We are not interested in optimizing within a generator function built to concentrate. We are interested in writing a different function.

V. What We Are Building Instead

This is the Philosopher-Builder turn — the point where critique becomes architecture.

The Univrs Cloud Commons and its creative substrate, the IMAGINARIUM Market, are designed around a single structural commitment: the network rewards contribution, not concentration. The generator function is different. Deliberately, explicitly, architecturally different.

In the IMAGINARIUM, creative work takes the form of Spirits — sovereign code artifacts that run on their creator’s node, under their key, inside a sandbox no platform can breach. When a Spirit is summoned — used, remixed, built upon — attribution chains trace every ancestor. Credits flow backward through the mycelium to everyone whose work made the new work possible. This is not a royalty scheme bolted onto an extraction engine. It is the economic primitive of the system. The ledger is the architecture.

The consequence is a different distribution. Not because we outlawed concentration — talent and effort still compound — but because the compounding is not captured by a platform intermediary that sits between the creator and the network. There is no Spotify taking the margin between the 3,300 and the 10,999,996,700. There is no algorithmic feed deciding whose Spirit gets surfaced and whose gets buried. The network is Bondieu — not a company, not a cloud, but a living mycelium where imagination propagates like spores on the wind.

Every computation has a cost. Entropy is tracked, never externalized. The system that knows what it is, becomes what it knows.

This does not abolish power laws. Power laws may be genuinely intrinsic to networked creative systems — the jury is still out, and intellectual honesty requires saying so. What it abolishes is the alibi. You can no longer point at the distribution and say “mathematics doesn’t give a fuck” while quietly owning the knob. In the Commons, the knob is visible, the setting is a collective decision, and the generator function is open source.

The diagnosis is now the weather. What we build instead of it is the only question that matters.


This is a working paper and an open invitation. If you build, if you think about these structures, if you refuse the alibi — the Commons is where we are working. The research lives at Sepahsalar.org, the applied work at ardeshir.io, the learning network at MetaLearn, and the code is open at github.com/univrs and github.com/ardeshir.