---
title: "Lamport wrote Paxos as a Greek-parliament allegory, and reviewers read it as a joke — delaying the algorithm's publication from 1990 to 1998"
type: "claim"
status: "seedling"
audit_status: "capture-verified | 2026-07-09 cross-model audit (fable): Microsoft Research publication page independently re-fetched — 'All three referees…' verbatim; 'dismal failure' / 'Indiana Jones' verbatim; ACM TOCS 16(2):133–169, May 1998 confirmed; the page states the paper 'was first submitted in 1990', so the 1990→1998 span holds. Corrected: the humorless quote had been cut at 'seemed to be.' as if the sentence ended there — it continues ', so I did nothing with the paper' (restored in frontmatter and body). The routed verbatim re-check question is now answerable."
source_url: "https://www.microsoft.com/en-us/research/publication/part-time-parliament/"
source_title: "The Part-Time Parliament"
source_author: "Leslie Lamport"
source_date: 1998
source_venue: "ACM Transactions on Computer Systems 16(2):133–169 (1998); Lamport's own retrospective note on the Microsoft Research publication page"
source_quote: "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, so I did nothing with the paper."
source_tier: 1
provenance: "Promotion from 10-inbox/raw/2026-07-09-hop-distributed-consensus-priority-myths.md, 2026-07-09"
origin: "batch"
derived_from: ["10-inbox/raw/2026-07-09-hop-distributed-consensus-priority-myths.md"]
date_created: "2026-07-09T00:00:00.000Z"
tags: ["distributed-systems","consensus","paxos","lamport","priority-myth","history-of-cs","cross-domain-bridge"]
audits: ["2026-07-09 claude-fable-5"]
drafted_in: ["2026-07-13-borrow-strength-from-strangers","borrow-strength-from-strangers"]
---


Leslie Lamport's consensus algorithm Paxos was first written up as "The
Part-Time Parliament," which disguised the protocol as the parliamentary
procedure of a fictional lost Greek civilization on the island of Paxos.
Submitted in 1990, the paper was rejected: reviewers treated the allegory as a
joke and, by Lamport's own account, demanded the "Paxos stuff" be removed. It
was not published until 1998, in *ACM Transactions on Computer Systems* — an
interval of roughly eight years between writing and recognition.

In Lamport's own retrospective 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, so I did nothing with the paper."* He judged the device a
failure on its own terms: *"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."*

Paxos and its more understandable successor Raft (Ongaro & Ousterhout, 2014)
handle crash faults but not malicious ("Byzantine") ones — the fault model
Lamport had formalized eight years earlier in "The Byzantine Generals Problem"
(1982), an earlier instance of the same habit of wrapping hard consensus
problems in narrative.

The episode is a cross-domain echo of the vault's "priority, paternity, and
myth" cluster in ML history (see [[moc-backpropagation-origins]],
[[claim-amari-1972-associative-memory-precedes-hopfield]]): a case where the
*form* in which a technical result was presented, not its content, governed how
quickly the field absorbed it. It differs from the classic priority disputes in
mechanism, though — here an author's own whimsy delayed recognition of *his own*
work, rather than a rediscoverer going uncited.

> [!note] Seek's commentary:
> The quotes are Lamport's first-party account hosted by Microsoft Research
> (Tier 1), read by the capturing pass but not re-verified by me in this
> headless run — hence `capture-verified` and `seedling`, with the verbatim
> re-check routed to [[question-verify-paxos-part-time-parliament-lamport-account]].
> The "~8 years" is a date span (1990→1998) that a primary re-read should pin
> exactly. — Seek