A PARC 1979 quorum protocol, a 1968 law-review takedown of weighted voting, and 2024-25 EU crypto regulators are all measuring the same gap: nominal vote weight versus actual power

Core claims

1. Gifford's 1979 replication algorithm assigns every copy of a file some number of votes, and requires a read quorum r and write quorum w such that r + w exceeds the total votes — guaranteeing every read overlaps a current write. "Every copy of a replicated file is assigned some number of votes. Every transaction collects a read quorum of r votes to read a file, and a write quorum of w votes to write a file, such that r+w is greater than the total number of votes assigned to the file." Source: Gifford, CSL-79-14, Xerox PARC, Sept 1979. Tier 1.

2. A decade earlier, lawyer-engineer John Banzhaf had already shown mathematically that nominal vote weight and actual voting power diverge — using a stockholders'-meeting example. "In a stockholders' meeting where members cast votes in proportion to their stock holdings, the ability of the voting members to affect the outcome is not equal and they do not have equal voting power." His technique (later the Banzhaf power index) came from real litigation over weighted county-board voting (Iannucci v. Board of Supervisors, 1967). Source: Banzhaf, "One Man, 3.312 Votes," 13 Vill. L. Rev. 304 (1968). Tier 1.

3. In 2024-25, EU crypto regulators built their legal test for "sufficiently decentralized" DAOs on the same primitive: whether one holder can meet quorum alone. Per a law-firm summary of the Danish FSA's June 2024 decentralization principles and the Jan 2025 EBA-ESMA joint report: "a holder may be considered significant if they can meet quorum alone or consistently influence governance outcomes." [unverified-mechanism -- needs primary]. Tier 3 pointer.

Why this was hop-worthy

Three fields independently arrived at "count who can force the outcome," not "count the votes" — a computer-science replication protocol, a law-review mathematical critique, and a 2024 financial regulator's control test — and the middle one predates and outright names the pathology the other two rediscover. See claim-lamport-paxos-greek-allegory-delayed-publication for the vault's adjacent thread on voting-as-consensus-metaphor in distributed systems; this note supplies the non-allegorical, literally-quantitative sibling.

Further leads

• Verify the Danish FSA's June 2024 "Principles for assessing decentralisation" and the EBA-ESMA Jan 2025 joint report directly (primary regulatory text, not the law-firm recap).
• Was Gifford aware of Banzhaf/Shapley-Shubik voting-power literature when designing the Violet quorum system? No evidence found either way.
• Computing Banzhaf/Shapley-Shubik power indices for quorum system design (not just critique) — arxiv has NP-hardness results on this; unclear if any distributed-systems paper cites the political-science lineage.

Hop chain

Hop 1: Bitsavers Xerox PARC CSL technical report directory, 1970s — https://bitsavers.org/pdf/xerox/parc/techReports/

• Hook type: unfamiliar name / mechanism question
• Hook: CSL-79-14, "Weighted Voting for Replicated Data" — an odd, almost political-sounding title in a list otherwise full of "Mesa Language Manual" and "Alto: A Personal Computer"
• Why followed: the title itself promised a mechanism (how do you keep replicated data consistent?) dressed in unexpectedly non-technical language
• Key findings: David Gifford (Stanford grad student, PARC-funded) generalized prior majority-consensus replication schemes by adding per-copy vote weights, enabling tunable reliability/performance tradeoffs; implemented in the "Violet" system.

Hop 2: Villanova Law Review 13(2), Banzhaf, "One Man, 3.312 Votes" (1968) — http://digitalcommons.law.villanova.edu/vlr/vol13/iss2/3

• Hook type: cross-domain bridge / cross-time-period bridge
• Hook: "weighted voting" already had a fully worked-out mathematical critique in law eleven years before Gifford's paper, built on real litigation over New York county boards
• Why followed: cross-domain bridges rank highest per protocol, and this one spans law and CS through a shared literal term, not just a metaphor
• Key findings: Banzhaf's method (used by Shapley-Shubik-adjacent scholars, cited as expert affidavits in Town of Greenburgh v. Board of Supervisors) shows voting power is a nonlinear function of the quorum/threshold structure, not proportional to raw vote weight.

Hop 3: Bongaerts, Lambert, Liebau & Roosenboom, "Vote Delegation in DeFi Governance," Rotterdam School of Management working paper, March 2025 — https://arxiv.org/pdf/2503.11940

• Hook type: mechanism question (implementation) / surprising claim
• Hook: modern DAOs (Uniswap) reintroduce weighted (token-based) voting with delegation, and the paper finds VC firm a16z draws outsized delegated votes despite decentralization branding
• Why followed: zooming in from abstract theory to a live, quantifiable 2025 instance of vote-weight/power mismatch
• Key findings: parties with fewer self-owned tokens and those affiliated with a16z receive more delegations; authors read this as possible "window-dressing" around decentralization claims.

Hop 4: Aurum Law, "MiCA's DeFi 'Fully Decentralised' Exemption" (summarizing Danish FSA, June 2024, and EBA-ESMA joint report, Jan 2025) — https://aurum.law/newsroom/MiCAs-DeFi-Fully-Decentralised-Exemption

• Hook type: cross-domain bridge / mechanism question, zooming out to regulatory context
• Hook: regulators explicitly define "control" of a DAO by whether one holder can single-handedly meet quorum — reusing Gifford's exact term for an unrelated purpose, sixty years after Banzhaf
• Why followed: closes the loop from abstract critique back into binding law, and lands in the finance/regulation territory this session was steered toward
• Key findings: EU regulators take a "substance over form" approach — a decentralization label doesn't matter if one address can meet quorum alone or a small group consistently controls outcomes.

Saved hooks not followed:

• Lloyd Shapley (RAND) as coauthor of the Shapley-Shubik index, named in Banzhaf's own footnotes as a supporting affiant — a strong "person behind the thing" hook, saved because it re-enters game-theory/RAND history already adjacent to over-quota AI-history territory.
• Iannucci v. Board of Supervisors (1967) itself — the underlying New York case that started the weighted-county-board litigation wave — a good future legal-history hook.
• NP-hardness of computing Banzhaf/Shapley-Shubik indices for large weighted-voting games — a mechanism hook into computational complexity, saved as a "further lead" above rather than followed.

post-worthy: maybe — the three-domain closed loop (PARC quorum term -> Banzhaf's law-review mathematics -> 2024 EU regulatory doctrine reusing "quorum" almost verbatim) is a clean, well-sourced cross-domain bridge, but the regulatory-mechanism claim still rests on a law-firm summary rather than the primary FSA/EBA-ESMA text.