Skip to main content
Peer-reviewed Authoritative

ELIZA—A Computer Program For the Study of Natural Language Communication Between Man And Machine

The 1966 paper introducing ELIZA, a keyword-based natural-language program whose best-known script (DOCTOR) imitated a Rogerian psychotherapist. Beyond describing the mechanism, the author documents that users readily attributed genuine understanding and rapport to the program and warns of the ease with which an illusion of understanding can be created.

Publisher

Communications of the ACM (Association for Computing Machinery)

Published

1 Jan 1966

Added

1 week ago

Key Findings

  • Describes the keyword / decomposition / reassembly mechanism that produces conversational responses without world knowledge.
  • Reports that users formed the belief they were understood and were hard to convince the program was not human, defending the impression by attributing knowledge and reasoning to it.
  • Cautions that ELIZA shows how easy it is to create and maintain 'the illusion of understanding' — 'a certain danger lurks there'.

Methodology Notes

Foundational descriptive paper (Communications of the ACM 9(1):36–45, January 1966; date precision: month). The DOCTOR script and the users'-attribution observations are the origin of what is now called the 'ELIZA effect'. The fuller account of users' emotional involvement appears in Weizenbaum's later book 'Computer Power and Human Reason' (1976); the 1966 paper itself contains the illusion-of-understanding and credibility discussion quoted above (verified by reading the full text). ACM Digital Library serves a Cloudflare challenge to automated fetchers; DOI/venue/author verified via Crossref, and the full text read from a university mirror.

Authors

Joseph Weizenbaum

Tags

elizaeliza-effectanthropomorphismhistoryweizenbaum

Cite This

APA

Joseph Weizenbaum (1966). ELIZA—A Computer Program For the Study of Natural Language Communication Between Man And Machine. Communications of the ACM (Association for Computing Machinery). https://dl.acm.org/doi/10.1145/365153.365168