Knowing Machines: Essays on Technical Change

by Donald Angus MacKenzie

Ranging from broad inquiries into the roles of economics and sociology in the explanation of technological change to an argument for the possibility of "uninventing" nuclear weapons, this selection of Donald MacKenzie's essays provides a solid introduction to the style and the substance of the sociology of technology. The essays are tied together by their explorations of connections (primarily among technology, society, and knowledge) and by their general focus on modern "high" technology. They also share an emphasis on the complexity of technological formation and fixation and on the role of belief (especially self-validating belief) in technological change. Two of the articles won major prizes on their original journal publication, and all but one date from 1991 or later. A substantial new introduction outlines the common themes underlying this body of work and places it in the context of recent debates in technology studies. Two conceptual essays are followed by seven empirical essays focusing on the laser gyroscopes that are central to modern aircraft navigation technology, supercomputers (with a particular emphasis on their use in the design of nuclear weapons), the application of mathematical proof in the design of computer systems, computer-related accidental deaths, and the nature of the knowledge that is needed to design a nuclear bomb.

Genres: Technology, Sociology

352 pages, Paperback

First published January 1, 1996

About the author

Donald Angus MacKenzie FBA FRSE FAcSS (b.1950) is a Professor of Sociology at the University of Edinburgh, Scotland.

His work constitutes a crucial contribution to the field of science and technology studies. He has also developed research in the field of social studies of finance. He has undertaken widely cited work on the history of statistics, eugenics, nuclear weapons, computing and finance, among other things. Works

Reviews

[David · Rating 3 out of 5]

Over lunch last summer, I was talking to a colleague about some work I'm involved with on micro-machined gyroscope design. I was very surprised by how much he knew about gyroscopes -- though he knows a lot of things, I didn't really expect him to come to the conversation knowing about Delco's HRG design, about the Sarnac effect, or about the trouble with lock-in in early laser gyroscopes. But then he told me that he'd read an interesting article by a "sociologist of technology" that talked about the design history of the laser gyro and the various influences that lead to its relative success. "I'll bring it for you to read," he said.

That was several months ago, and this evening I finally sat down to read it. It is indeed a very interesting book. Apart from the chapter on the Sagnac effect and the evolution of the laser gyro, I was intrigued by the essay on how the national labs influenced supercomputing (something rather pertinent to me), and the essay "The Charismatic Engineer" about Seymour Cray. And then there's the essay "Negating Arithmetic, Constructing Proof." The first half of which was about the development of the IEEE 754 floating point standard, a topic that I happen to know in intimate detail, and the history of which I know quite well -- and never thought that I would read about from the perspective of a sociologist!

The most interesting essay, though, was the last one: "The Uninvention of Nuclear Weapons." As a computational scientist who works closely with experimentalists, this was, again, almost unsettlingly close to home. The main point of the article was that nuclear weapon design involves a lot of tacit knowledge (what one friend calls "engineering sense") about the limits of different simulation codes and approximation methods; and that tacit knowledge is built on comparison of those codes and approximations to real experimental data. With a nuclear test ban in place, will that tacit knowledge fade, thus possibly "uninventing" the practical aspects of nuclear weapon design? Even granting that the world might have been better off without nukes, I find this thought somewhat unsettling. What more benign tacit knowledge might we be losing even now?

[Nick Black · Rating 3 out of 5]

Amazon 2009-08-05, impulse acquisition. This book suffered from being written by a sociologist, who (1) wasted a bunch of time up front with extended (about three chapters' worth) nonsense of a sociological nature (i continue to wait eagerly for something of merit to emerge from this "social science") and (2) terminated each of the subsequent admittedly-pretty-good essays on technological development with a confusing, unsettling reference back to the first sociological principles detailed early on. On the other hand, this pattern became useful, because it signaled precisely where one ought stop reading each chapter. Most laughable was his suggestion that we can "uninvent" nuclear weapons -- brother, I'm a contradiction of that hypothesis all my own. I'm rather offended at the ethical aspects of such a suggestion, perhaps more than the stupidity -- yes, Mr. MacKenzie, let's please return to an era where macroscopic, protracted war via chemical and mechanical death every few decades is the norm. Let's look to Rwanda, Sudan and Congo as fine examples of the base decency of man when the nuclear genie's still in the bottle. I'd far rather be hacked to death with a machete by my neighbor than leave some megatonnage under PAL locks! Fool, you have nuclear weapons to thank.

[Arkuat · Rating 4 out of 5]

This basically qualifies as sociology of technology, which is a little bit like sociology of science, so if you're allergic to that stuff, you've been warned.

[Ken-ichi]

Not really sure I want to read all of this, but I remember reading some of them in school and they were interesting. Uninvention is an interest of mine.