A 1996 book argued people apply social rules to computers

The design

In 1996, Stanford communication researchers Byron Reeves and Clifford Nass published The Media Equation: How People Treat Computers, Television, and New Media Like Real People and Places, jointly issued by CSLI Publications and Cambridge University Press. The publisher's own catalog record lists the book at 305 pages, filed under Library of Congress subject headings for mass-media audiences, media influence and the psychological effects of media. The Internet Archive's library record for the same edition confirms the 1996 date and the joint imprint. Neither record reviews the book's arguments; each documents only that the book exists, who wrote it, and how a library classifies it.

What the evidence says

This entry was not built from a chapter-by-chapter reading of the book, which is not available in an openly readable full text; it relies on the book's own catalog record, plus the peer-reviewed paper the same lead author, Clifford Nass, had already published two years earlier at the CHI 1994 conference, covered separately by this site. That 1994 paper, recorded in ACM's own Digital Library entry, named the underlying claim later expanded into the book: that people apply social rules from human relationships to computers and other media. A catalog record cannot confirm every experiment in the book, but it establishes, independently of publisher marketing language, that the book reports an experimental research program rather than a single case study of one product.

What it asks of people

A book cataloged under mass-media psychology asks nothing of a reader directly; it is a scholarly work, not a product with terms of use. What matters is what later writing does with its claim: treating 'people respond socially to media' as a finding from controlled studies conducted in the 1990s differs from treating it as a statement about any present-day companion chatbot, which the authors never tested. Writing that cites the 'media equation' as a settled explanation for attachment to a modern AI companion extends a decades-old, product-agnostic finding, and that extension belongs to the later writer, not the book.

Privacy and safeguards

Neither the catalog record nor the library metadata addresses data handling, because the book is not a software product and collects no user data. The record does supply a documented boundary: this is academic research into television, computers and early media conducted before generative AI existed, and any safeguard argument built on it needs its own, separate basis for a modern companion product.

• Does a citation of the 'media equation' name which experiment it relies on, or only the book's title?
• Is the claim being extended to a product the original authors could not have tested in 1996?
• What would a reader need to read in the book, rather than a summary, to judge whether its findings still generalize?

The publication record fixes the book to a year, a publisher and an academic lineage; it does not resolve how far its claims should be read into contemporary companion AI, and that gap is worth naming whenever 'media equation' appears in later writing about chatbot attachment.