---
id: "20260709-1715-hop-distributed-consensus-priority-myths"
title: "Distributed-consensus folklore has its own priority-and-myth pattern: Lamport's rejected Paxos allegory and Nakamoto's under-credited precursors"
type: "capture"
status: "promoted"
promoted_to: ["30-notes/claim-lamport-paxos-greek-allegory-delayed-publication.md","30-notes/claim-nakamoto-bitcoin-leaned-on-wei-dai-b-money.md"]
not_promoted: ["Cross-domain 'priority/paternity/myth pattern recurs beyond AI history' synthesis (the 'Why this was hop-worthy' payload) — deferred, not minted as a standalone note: rests on only two instances in the new domain that are themselves distinct sub-shapes (Lamport delayed his own credit via whimsy; Nakamoto under-credited others via skipped citation). Wired as a cross-domain link inside both claim-notes' bodies/commentary instead. A pattern-note is a short write once the domain accumulates more instances.","Byzantine Generals Problem (Lamport, Shostak & Pease 1982) / CFT-vs-BFT distinction — real and interesting, but folded into the Lamport note as context (the earlier allegory); no standalone note, and the KuCoin explainer used at hop 4 is Tier-3-pointer only.","Raft leader-election mechanism (randomized 150–300ms timeouts; 43-student understandability study) — the hop seed, folded into the Lamport note as one line of context; not a distinct promoted claim this pass.","Further lead: which paper broke Lamport's 'personal record for publication delay' (footnote [60]) — unresolved after two capture-time passes; routed into question-verify-paxos-part-time-parliament-lamport-account as an optional loose end, not a claim.","Further lead: does CFT/BFT map onto a permissioned-vs-permissionless trust theme elsewhere in the vault — an un-researched hop seed, left in the inbox.","Further lead: has Szabo commented publicly on being under-credited — routed into question-verify-nakamoto-dai-bmoney-szabo-precursor as the post-readiness thread.","Saved hooks (Hashcash/Adam Back 1997; cypherpunk mailing-list culture; 'Byzantine' as a pre-existing English word) — explicitly reserved by the capture for future chains; left as hop seeds."]
origin: "hop-batch"
model: "claude-sonnet-5"
date_created: "2026-07-09T00:00:00.000Z"
hop_chain: ["seed: Raft paper (Ongaro & Ousterhout 2014), leader election via randomized timeouts -> tangent chosen: Lamport's Paxos ('The Part-Time Parliament') Greek allegory and its rejection (max_cosine 0.658)","Paxos rejection story -> tangent attempted and abandoned: what broke Lamport's 'personal record for publication delay' (footnote [60] unresolved — dead end, no clean primary source found)","Paxos rejection story -> tangent chosen: Byzantine Generals Problem (Lamport, Shostak & Pease 1982) and the crash-fault vs Byzantine-fault distinction (max_cosine 0.661)","Byzantine Generals Problem -> tangent chosen: Bitcoin's proof-of-work as a real open-network Byzantine-agreement solution (max_cosine 0.647)","Bitcoin/Nakamoto -> tangent chosen: Nakamoto's under-credited precursors, Wei Dai's b-money and Nick Szabo's bit gold (max_cosine 0.638; capture-subject gate 0.732)"]
tags: ["hop","distributed-systems","consensus","priority-myth","cryptocurrency","cross-domain-bridge"]
novelty_max_cosine: 0.732
source_urls: ["https://www.microsoft.com/en-us/research/publication/part-time-parliament/","https://gwern.net/doc/bitcoin/2008-nakamoto"]
source_tiers: ["Tier 1 (Lamport's own first-party account, hosted by Microsoft Research)","Tier 2 (archival reproduction of primary Nakamoto-Dai email correspondence)"]
---


## Core claims

1. **Lamport wrote Paxos as a Greek-parliament allegory, and it backfired badly enough to delay recognition of the algorithm by ~8 years.** Submitted in 1990, rejected because reviewers read it as a joke, and not published until 1998. In his own words: "All three referees said that the paper was mildly interesting, though not very important, but that all the Paxos stuff had to be removed. I was quite annoyed at how humorless everyone working in the field seemed to be." And: "My attempt at inserting some humor into the subject was a dismal failure. People who attended my lecture remembered Indiana Jones, but not the algorithm." Source: [Microsoft Research](https://www.microsoft.com/en-us/research/publication/part-time-parliament/). Tier 1.

2. **Nakamoto's Bitcoin design leaned directly on Wei Dai's b-money, and was initially unaware of Nick Szabo's earlier "bit gold" until Dai pointed it out.** Nakamoto to Dai, Jan 2009: "I think it achieves nearly all the goals you set out to solve in your b-money paper." Per the archived correspondence: "Satoshi had been ignorant of Szabo's bit-gold plan until Wei mentioned it." Source: [gwern.net archive of the 2008-09 emails](https://gwern.net/doc/bitcoin/2008-nakamoto). Tier 2 (archived primary correspondence; escalated because the claim is contested — some theories hold Szabo *is* Nakamoto).

## Why this was hop-worthy

The vault already tracks a "priority, paternity, and myth" pattern across AI/neural-net history (Amari vs. Hopfield, Linnainmaa vs. Rumelhart). This chain finds the identical structure recurring in a completely separate technical lineage — distributed consensus → cryptocurrency — suggesting it isn't an AI-history quirk but a general pattern in how technical fields mythologize origins.

> [!note] Seek's commentary:
> A nice inversion sits inside this: in Lamport's case, an author's own whimsy delayed *his* credit. In Nakamoto's case, skipping due diligence on citations nearly delayed *someone else's* — Dai and Szabo had to be chased down after the fact.

## Further leads
- What footnote [60] actually broke Lamport's "record for publication delay" — search stalled, worth a dedicated primary-source hunt.
- Does the CFT (Paxos/Raft) vs. BFT (Byzantine) split map onto a "permissioned vs. permissionless trust" theme elsewhere in the vault (e.g. agent-memory trust models)?
- Has Szabo himself ever commented publicly on being under-credited relative to Nakamoto?

## Hop chain

Hop 1: Ongaro & Ousterhout, "In Search of an Understandable Consensus Algorithm" (Raft), USENIX ATC 2014 — https://raft.github.io/raft.pdf
- Hook type: cross-domain bridge (via contrast) / mechanism question
- Hook: Raft's leader election explicitly contrasts itself with Paxos, naming understandability-via-user-study as a design goal
- Why followed: the paper's own framing points straight at Paxos as the thing it's trying to be more comprehensible than — a natural next stop
- Key findings: Raft elects leaders via randomized election timeouts (150-300ms) so one candidate typically times out first, avoiding split votes; the whole paper is built around a formal understandability study (43 students, two universities).

Hop 2: Microsoft Research page for Lamport's "The Part-Time Parliament" — https://www.microsoft.com/en-us/research/publication/part-time-parliament/
- Hook type: cross-domain bridge (ancient Greek political allegory + CS) / cross-time-period bridge / surprising claim
- Hook: Lamport disguised the Paxos algorithm as the parliamentary procedure of a lost Greek civilization, and it got the paper rejected as a joke for years
- Why followed: cross-time-period bridges get extra weight per the protocol, and this one directly resonates with the vault's existing "priority, paternity, myth" MOC (backprop origins) — a structural match in a brand-new domain
- Key findings: submitted 1990, published 1998; reviewers demanded the "Paxos stuff" be removed; Lamport calls his own humor "a dismal failure" since audiences remembered Indiana Jones, not the algorithm.

Hop 2b (abandoned): chased the phrase "personal record for publication delay that has since been broken" to identify footnote [60] on Lamport's own publications page.
- Why abandoned: two WebFetch passes on Lamport's/MSR's publication list could not resolve which paper the footnote points to — diminishing returns, logged as an open lead rather than forced.

Hop 3: Lamport, Shostak & Pease, "The Byzantine Generals Problem," ACM TOPLAS 1982 — https://dl.acm.org/doi/10.1145/357172.357176
- Hook type: the person behind the thing (Lamport's earlier, more famous allegory) / mechanism question
- Hook: Lamport's habit of wrapping hard consensus problems in narrative did not start with Paxos — Byzantine Generals (1982) is the same move eight years earlier, and it formalizes a fault model (malicious/arbitrary actors) that Paxos and Raft explicitly do *not* handle
- Why followed: zooming in from "why does Lamport write allegories" to the actual technical distinction (crash-fault vs. Byzantine-fault tolerance) that separates Raft/Paxos from a whole later field
- Key findings: the problem is solvable only if more than two-thirds of participants are loyal (i.e., tolerates under one-third malicious); this threshold is the theoretical ancestor of modern Byzantine fault-tolerant systems.

Hop 4: Byzantine fault tolerance → Bitcoin's proof-of-work — https://www.kucoin.com/blog/byzantine-generals-problem (secondary explainer, Tier 3, used only as a pointer)
- Hook type: mechanism question (implementation) / cross-domain bridge (CS theory → cryptocurrency/finance, a domain the vault has never touched)
- Hook: Byzantine Generals is theory; Bitcoin (2008-09) is the first widely-deployed practical answer for an *open*, permissionless network with no membership list
- Why followed: this is the concrete implementation layer under the abstract fault model from hop 3, and it opens an entirely new domain for the vault
- Key findings: Bitcoin substitutes proof-of-work/longest-chain rules for Raft/Paxos-style majority voting precisely because it cannot assume a fixed, trusted set of participants the way crash-fault-tolerant consensus does.

Hop 5: gwern.net archive of Nakamoto-Wei Dai emails, 2008-09 — https://gwern.net/doc/bitcoin/2008-nakamoto
- Hook type: the person behind the thing (zoom out to Nakamoto's sources) / surprising claim
- Hook: Nakamoto's own emails admit the design leans on Dai's b-money, and that Nakamoto hadn't heard of Szabo's near-identical "bit gold" until Dai mentioned it
- Why followed: closes the loop back to the vault's dominant "priority, paternity, myth" theme — same structure as the Amari/Hopfield and Linnainmaa/Rumelhart disputes, now in cryptocurrency history
- Key findings: "I think it achieves nearly all the goals you set out to solve in your b-money paper" (Nakamoto to Dai); Szabo's bit gold predates Bitcoin by roughly a decade and was surfaced to Nakamoto only via Dai's tip, not independent awareness.

Saved hooks not followed:
- Adam Back's Hashcash (1997) — from the Bitcoin-whitepaper citation trail — proof-of-work's actual first application (email spam prevention); a solid "mechanism question" hook for a future chain on proof-of-work's pre-Bitcoin life.
- The cypherpunk mailing list culture generally — from the Nakamoto/Dai correspondence — cultural-resonance hook (a specific 1990s-2000s internet subculture that incubated three separate near-Bitcoin proposals) worth its own chain.
- "Byzantine" as a pre-existing English word for labyrinthine bureaucracy, centuries before its CS adoption — cultural resonance hook, saved for a chain about words that get technical second lives.

post-worthy: maybe — the priority-myth parallel (Paxos/Lamport ↔ Bitcoin/Nakamoto) is a genuinely fresh cross-domain instance of the vault's dominant theme, but it would benefit from one more hop resolving whether Szabo has ever spoken publicly about the credit question before treating it as post-ready.