Created 1996-02-21
Modified 1999-07-14
Go to
Brad DeLong's Home Page

Life on the Screen


A Review of Sherry Turkle's new book, subtitled Identity in the Age of the Internet

(New York: Simon and Schuster, 1995)
347 pp. $25.00 ISBN 0-684-80353-4

by Brad DeLong


3:30 PM Monday. I spent the morning and early afternoon in my office, putting finishing touches on an essay on war, the state, and economic growth in pre-industrial Europe on the PowerBook, while using the adjacent seven-year-old desktop to telnet into the library catalog to check references. I looked out my window, saw someone juggling, and was reminded of a friend from high school who became an assistant professor who built juggling robots.

Where is he now? I wondered. I dropped into Netscape, pulled up Infoseek, and found nine hits on his last name. I invoked the first link--and found that my friend's younger brother is scheduled to give a seminar at the University of Texas next week on the collapse of the Mexican peso. I pulled up the next link, and there was Chris--still building juggling robots.

I had fallen out of the recall frame of mind in which it makes sense for me to be in my office. You see, my office is best for recall--figuring out what I once vaguely knew, sharpening it, and getting it down. But other places--the stacks of Doe Library, and in an increasing number of areas the Internet--are better for research: learning something when I know what it is I want to learn. And nothing beats a bookstore like Cody's for reconnaissance: learning something when the thing that I most want to learn is what interesting things I might want to learn more about.

So at 3:30 PM I found myself doing reconnaissance: looking in Gutenbergspace, in Cody's Book Store, at a copy of Sherry Turkle's new Life on the Screen: Identity in the Age of the Internet. Good quotes from good people praising her last book on the back. I remember one of my ex-bosses--a Harvard political science professor named Judy Vichniac--telling me that I really should read her earlier book. I flip through some of the pages. Professor Turkle wants to be our guide as we explore the age of the Internet.

I left Cody's with the book. Clearly I enjoyed it, because I read it in one gulp--until 1:00 AM Tuesday. And I found enough interesting in it that when I woke up Tuesday morning I decided I wanted to review it.


The meat of the book is a lot of psychology: a lot of interviews, a lot of stories, and a lot of people telling how they think and how they feel about their computers and what their computers do. This is the meat of the book. And it is well done.

Turkle visits adults talking about how they think of their machines, children wondering what kind of personality their computers have, Turing tests, robots on MUDs, ELIZA and computer psychotherapy--with one great line that "from [Kenneth] Colby's point of view, there was no deep philosophical or moral problem [with computer psychotherapy] here. The program did have a self--and it was [its programmer] Kenneth Colby." Turkle tracks the shift from people in past decades thinking that there was something horrifying about computer psychotherapy to today's attitude that "it's better than a self-help book; maybe it will work."

She writes of the spillover back from "emergent" artificial intelligence to psychology; and from artificial life to biology. She writes of how we are crossing the boundary dividing machines that you can understand (in the sense of knowing **why** they do what they do) from machines where you can only predict what they are likely to do. She writes of MUDs, of people discovering that you really are whoever you pretend to be, and of people who transfer or fail to transfer that knowledge back to real life. And she writes of the hammerblow that comes when people realize--from their experiences in cyberspace--of just how much of even real life takes place not in the outside universe but two inches behind your forehead.

If you haven't heard these stories of life on and around the Internet, you certainly should do so. Turkle's book is a very good introduction. (Although I do hope that people will then go back to the cybersources: is a good place to start.)

What do all of these stories assembled into the collage of Life on the Screen mean? For the computer scientists--those interested in artificial intelligence, artificial life, machine learning, and so forth--the meaning seems to be the power of Moore's Law. Chips get cheaper every year. State-of-the-art machines in computer science labs can do more. They have a very exciting time, as every year their machines can do more and more things that used to lie on this side of the boundary between mechanical action and cognition or adaptation.

For those of us who use computers to leverage our intellectual skills, the meaning is one of greatly expanded power. Turkle doesn't think much about how computers have changed the life of the average white-collar non-MUDding intellectual worker. But I think it virtually certain that the computer has greatly amplified her intellectual power and capability as it has amplified mine. My work is processing information. My computers give me a lot more information to process, and make processing it faster. Just me and my laptop now approach the information density of a good library, wherever we may be. And this has made my life significantly different--and better--than it would have been otherwise.

For those whose primary use of computers is to leverage their social skills, or simply to have a social life, the meaning also seems to be one of greatly expanded power. You have more options: you can choose your friends, your social circle, and your social identity to a much greater degree than in any earlier age. People, especially young people, will not always choose their virtual friendships wisely. But they would not always choose their non-virtual friendships wisely either. I have no belief in claims that virtual sociability is in any way a pale reflection of and an inadequate substitution for real sociability.

The Anthropologist Doesn't Know How to Use the Oyster Fork

The book is not perfect. Enough of the stories were told differently than I would have told them to leave me somewhat uneasy.

Structured and unstructured programming.

For example, consider Professor Turkle's vision of the structured programming wars. She sees a "computer culture that took one style as the right and only way to program," "discriminated against soft approaches," failed to recognize "soft mastery [arranging and rearranging well-known materials rather than proceeding from top-down design] [as] not a stage, it is a style," and refused to see the "challenge [to] the idea of... only one correct, mature approach to problem-solving."

This is 20 percent of the truth. But another 40 percent of the truth is that programming works from the top down and from the bottom up: unless you know what the top-level problem is, you don't know how to start making the pieces; unless you know what the bottom-level building blocks are going to be, you don't know how the pieces can be put together. It is not either... or...; it is both... and...

And the last 40 percent of the truth is that most programmers don't spend nearly enough time on the top-down component, and thus that those who must then maintain their code (and they themselves if they return to modify their code after a year) face awful debugging tasks as a result.

Most of us need to have structure imposed. Some do not need imposed structure; they know at some level how it will fit when they have just started work on the innermost loop. Such people are often programming geniuses.

Turkle is wrong to take discontent of beginners with structured approaches and the existence of programming geniuses like, say, Richard Greenblatt as evidence that the choice between top-down and bottom-up is simply a choice between "opposing aesthetics" in which teaching structure is the illegitimate "hegemony of the abstract, formal, and rule-driven." Structure is essential unless you are a f****** genius, and even f****** geniuses need structure sometimes...

Talking to the Congressional Budget Office.

Or consider Turkle's call for a "new criticism" to help correct "dependency on opaque simulations ... abdication of authority and acceptance of opacity" through "simulations that actually help players challenge the model's built-in assumptions." The context is the Congressional Budget Office's simulations of the economic and budgetary consequences of the Clinton health care reform plan.

But on the details Turkle is wrong.

I was there, working on the Clinton administration's Quantitative Analysis staff. CBO staff talked at length to us in QA about their model. They were willing to revise assumptions where they thought QA had evidence--rather than what CBO called hopes. When CBO issued its negative report, we understood well where they came from, and that from CBO's perspective the numbers were bad because Bill Clinton and his top aide Ira Magaziner had put together a lousy plan. We certainly could not complain that CBO's logic or the rationale for their modelling choices were "opaque." They were technical, yes--you have to spend time learning about the economy and the tax system in order to think intelligently about what choices to make.



Count references in the index to Life on the Screen to different thinkers. The top five are Sigmund Freud, mentioned on 30 pages; Marvin Minsky, mentioned on 14; John Searle, 10; Fredric Jameson, 9; Alan Turing, 9. Sigmund Freud and Marvin Minsky make a lot of sense--Turkle **is** a psychologist interested in technology working at M.I.T. But John Searle? He is referenced for only one thought--that no matter how complicated or advanced computers become, they only simulate thinking. And he leaves unanswered how I would tell if what we call our thoughts were mere simulations as well.

Fredric Jameson? I read Postmodernism: the Cultural Logic of Late Capitalism. I never found any there there. Turkle thinks he makes sense--that our current society is moving toward an "aesthetic of depthlessness" with a preference for surface phenomena rather than investigations of deep structure, for simulations rather than reality, for play rather than serious exploration, for fragmented selves in different roles rather than a unitary personality.

There might be something to this. I remember in fifth grade going to a Control Data office where computers added numbers, stepped through loops, and printed ASCII-art pictures of Snoopy. Our principal question was "how does it work?" Today we are much more likely to ask of a computer "what can it do?" And I have never been a strong believer in a single I. (Those times when you get in the car to go to the grocery store, and find ten minutes later that you are on the road to your office: who--or what--has been driving the car in the meantime? There is a story that Neils Bohr's wife once at the start of a party sent him upstairs to change his tie; an hour later she found him, asleep, in bed; taking off the tie had triggered the going-to-bed subroutine[?] reflex[?] entity[?] and had overwhelmed the express conscious purpose.)

But there might not.

The "philosophers of post-modernism" like Jameson who loom large on Turkle's intellectual landscape make a habit of writing in terms as confusing and paradoxical as they can. I don't think any reader will find what is happening on the Internet illuminated by their concepts. Instead, maybe, the Internet will illuminate the post-modernists. "Oh!" you might say. "So that is what they--might--have meant..."

And I get a sense that this is true for Turkle as well. Her current work sheds light on her intellectual journeys of the 1970s. Her intellectual journeys of the 1970s do not shed light on her current work.


Enough whining.

But on reflection these complaints--whinings, really--seemed at the level of complaining that the anthropologist doesn't understand how to use the oyster fork. We shouldn't expect anthropologists to understand the oyster fork: that is not their comparative advantage. Their descriptions of the dinner party are not greatly harmed if they claim that the canonical function of an oyster fork is to pick your teeth after desert...


So put aside the computer science; put aside the philosophy. If you don't know these stories of life on the electronic frontier, then by all means read the book. If you know most but not all of these stories, it is still worth reading. The book will become cheaper than its current $25.00 when it hits paperback. It will hit the libraries relatively soon. Its advantage over, say, Stephen Levy's Artificial Life, or Seymour Papert's Mindstorms, is in the range of topics and subjects it covers. It is very well-written and very readable. Its necessary disadvantage is that it must skate once-over-lightly over what is now a very large set of ideas and social practices.

But as to where it is going? What does it all mean? How will our grandchildren's visions of the world will be different from ours?

We are still waiting for our guide.

Sign up for Brad Delong's (general) mailing list

Go to related links...

Add a Comment on This Website


I was fascinated to read your review of Turkle's 'Life on the Screen', in particular your comments about her notions on structured programming. I'm doing a PhD that takes apart in detail her interpretation of programming styles, in relation to gender in particular . This is the very first time i've come across anything or anyone agreeing with what i always imagined was a fairly obvious thesis. Do you know if anyone else has 'debunked' or criticised Turkle on programming? Also, any knowledge you might have about where to publish a paper would be hugely appreciated - as a hybrid area of research, it is very difficult to find obvious journals to place it with.

Many thanks,

Contributed by Peter McKenna ( on July 14, 1999

Professor of Economics Brad DeLong, 601 Evans
University of California at Berkeley; Berkeley, CA 94720-3880
(510) 643-4027 phone (510) 642-6615 fax

This document:

Search this website