u_topn picture
rub_dc picture
Atlantic Unbound Sidebar

Discuss this topic in the Digital Culture forum of Post & Riposte.

Previously in Digital Culture:

"Caught in the Flash," by Harvey Blume (December 1997)
Computer memory, once thought to be finite, now looks like an open frontier. The author peers into the future of memory and sees a paradox rooted deep in our past.

"Dispelled," by Ralph Lombreglia (November 1997)
Why Riven -- the most anxiously awaited multimedia product ever -- is missing the magic of Myst.

"God, Man, and the Interface," by Harvey Blume (October 1997)
Most of us take the computer interface for granted. But for Steven Johnson it is a defining metaphor of our times -- and a summons to the metaphysical.

For more, see the complete Digital Culture Index
The World According to David Gelernter
An interview with a computer scientist who argues that beautiful technology -- and a return to traditional values -- must show us the way forward

by Harvey Blume

January 29, 1998

Ambivalence is an underlying -- and underdocumented -- feature of the computer revolution. If you're neither a digital billionaire nor Nicholas Negroponte (not that the two are mutually exclusive), you are likely to entertain doubts about computers -- even as you lend your intelligence to them. Scratch a Webmaster, and you may hear worries about loss of privacy. Have a beer or two with a video-game designer, and don't be surprised if he starts crabbing about how we are programming ourselves into digital dystopia.

The Yale computer scientist David Gelernter entertains just about as much ambivalence toward computerization as it is possible to bear without succumbing to paralysis. In his epilogue to Mirror Worlds: or the Day Software Puts the Universe in a Shoebox ... How It Will Happen and What It Will Mean (1991), Gelernter divides himself into a humanist and a technologist who ramble together through New Haven's byways arguing about the onrushing virtualization of experience. The outcome of the debate looks too close to call until the technologist comes out with this knockout of a rejoinder to the claim that technology lacks feeling: "Transformed childhood joy. That's why we do technology . . . It's all emotion. When you think of technology, that's what you ought to think of. The kid riding his bike, or sledding downhill, or charging over a grass field trying to get his kite to fly, just because it feels great, it's the human thing to do."

In Gelernter's The Muse in the Machine: Computerizing the Poetry of Human Thought (1994) -- a meditation on artificial intelligence, dreams, creativity, and difficult passages of the Hebrew Bible -- the humanist gets revenge. To the grand ambition of computer science, the desire to one day build a machine with a mind every bit as conscious as a human mind, this book supplies a sharp rebuff. The computer can be programmed to imitate many aspects of consciousness, but software, however useful, can never endow a machine with a mind of its own. In Gelernter's words: "We might ultimately build a computer that seems to us to have a mind. But I doubt whether the computer itself will ever be taken in."

Gelernter's latest book, Machine Beauty: Elegance and the Heart of Technology, just published this month, attempts to build a bridge between humanism and technology, art and science. Our sense of beauty is that bridge. Beauty is a defining element not only of great poetry and great mathematics but also of efficient machines and elegant software. When it comes to software, beauty, in Gelernter's view, is not a matter of decoration, not an add-on -- it is indispensable. Since software is bound by relatively few physical restraints, only the demand that it be beautiful will keep it from becoming dangerously complex. If cultivating a feeling for beauty implies changing the way we educate computer scientists, so be it, says Gelernter; art and art appreciation are essential to the new technology.

Drawing Life Gelernter is a vivid, engaging writer, and it may well be his literary qualities -- rather than his achievements in parallel processing or interface design -- that brought him to the attention of the Unabomber. In any event, when Gelernter opened a package in his office early one morning in June, 1993, it turned out to be an explosive, obviously built to kill -- and it nearly did. Gelernter's attempts in Drawing Life: Surviving the Unabomber (1997) to make sense of the media's and society's responses to the bombing differ from his other work in that his customary talent for wresting clarity from ambivalence is replaced by a tone of right-wing stridency. But there's no doubt that David Gelernter would like these political views to be taken as seriously as his others, not discounted as the consequences of an encounter with terrorism.

Conversation with David Gelernter moves easily from art to engineering, science to politics -- as it did in this discussion, which took place during the turmoil of Theodore Kaczynski's trial.



In Machine Beauty you write, "the scientific and artistic personalities overlap more than they differ." What role do computers have in bringing art and science together?
See a series of excerpts from David Gelernter's Machine Beauty: Elegance and the Heart of Technology (Basic Books, 1998).
I want to make this claim about art and science overlapping completely independently of computers. It's a characteristic of human nature we see in classical times, and certainly in the Renaissance. We tend to think of scientists who look at art, and artists who look at science, as aberrations, but there are a lot of aberrations, and it is often the most talented people who manifest the aberration.

Would you say there is an increasing amount of discussion these days about the opening of the borders between art and science?

"Beauty is a truth-and-rightness meter, and science and technology could not exist without it."

Machine Beauty In a way I think that's true. There's more such discussion today than there was five or ten years ago, and that's desirable. But what's going on is zero compared to what was going on in 1930 or 1940. The two communities were vastly closer than they are today. The schools were vastly more serious then. Public schools were generally respected and generally did a good job. Colleges, particularly public colleges, were getting more serious. It went without saying that if you were getting a B.S. and were a smart young scientist, you were going to sit in on an art-history survey course, you were going to know something about art.

Every college graduate in the thirties and forties was talking about Picasso; it was on the public agenda. If you read Alan Turing's famous 1950 paper about the Turing test, you find it's natural for him to throw in a reference to Picasso.

You write that Turing "fit into the ambient intellectual climate" of his day in his attempt to "strip computing to its basics and seek the simplest possible all-purpose computing machine." You compare him to Russell and Whitehead in their striving to get at the roots of mathematics. If so, wasn't he also like Freud, Stravinsky, and Picasso in their efforts to get to the essence of psyche, music, art?

"Society's reaction to machine beauty brings Freudian resistance to mind. Metaphorically, I mean, but the metaphor is too apt to pass over."






"Insisting that beauty is at the heart of science and technology is like ordering wine at lunch, or tacking ruffles to your office furniture."

Absolutely right. In the case of Picasso, it's explicit. He wanted to go back to the primal and original. Stravinksy had exactly the same goal -- as did Freud, in trying to determine the origins of modern consciousness. There are echoes of Freud in Turing. When Turing says, for example, that he can't answer the religious objection to his proposal with regard to machine thinking because it is really a demand for consolation, this is an explicit echo of the conclusion to Freud's Civilization and Its Discontents. Again, it goes without saying that educated scientists and mathematicians like Turing knew the work of Freud and Picasso, not because these people were polymaths but because that's what education meant.

Where does science end and art begin? How do you separate the two?

It's a good question. It's probably more a question of what your materials are than what your mind is doing. I'm not convinced that the character of mental activity is all that different. Some artists are drawn to paint and some are drawn to marble --

And the ones drawn to subatomic particles are physicists?

Exactly. It's a question of medium more than mentality.

Many computer scientists claim we will one day be able to build machines that are intelligent in the fullest sense. From what I gather, you are not too optimistic about the prospect.

There's a very important distinction here. Can computers compete with humans in the solution of delicate and complex tasks? The answer is yes. Does that mean they are becoming human-like and acquiring minds? The answer is definitely no.

That can't happen?

In principle it can't happen. There is a vocal group of true believers who think that mind can evolve out of software, but I think they are considerably fewer than they used to be. I don't think I'm as far outside the consensus as some people hold me to be when I say the arguments against machine intelligence are not only right but obviously right.

These days it seems theories of mind and theories of computation can't help but have deep implications for each other.

I published a book, The Muse in the Machine, about this question. You cannot have a theory of the human mind, as computer people are trying to do, that isn't also a theory of hallucinations, a theory of dreaming, and also a theory of nightmares. I do believe, on the other hand, that you can have a theory of mind that says nothing at all about computers.

Where's the firewall? What is it that a mind does that a computer can't?

The brain is not a computation machine. It's a machine with certain extraordinary properties -- system properties, ensemble properties, not properties of particular neurons but of the way the whole thing is put together. We know it is a property of the brain to create consciousness, to generate an "I", raw feelings, sensations, mental states. How it creates these mental states, we don't know for sure.

Take an utterly different machine: a computer made of silicon, plastic, and electronic circuitry. What gives me the right to believe this radically different machine made of different stuff arranged in a radically different way can do this amazing trick of creating consciousness? I simply have no reason to think that it could -- even if I can't prove that it can't.

"Beauty is more important in computing than anywhere else in technology.... As we throw off the limits, what guides us? How do we know where to head? Beauty is the best guide we have."

We are using machines to do more and more of what we used to do. There is both enthusiasm and unease at the prospect, which your writing reflects very well.

The enthusiasm and the unease are both fully justified. There is a sense in which software can't take over culture without hugely diminishing it. There was, a year or two ago, a thing called the Internet World Exposition -- a World's Fair on the Internet -- and a lot of people asked me about it because I'd just published a book about the New York World's Fair of 1939 [1939: The Lost World of the Fair (1995)]. I said it's a fundamental misconstruction of what a real World's Fair is to believe one could do it in cyberspace. People went to a World's Fair precisely to be with other people, see the sights, smell the smells, eat the hot dogs, see and hear the flags waving. Replacing physical experience with sterile images on a computer screen does not add up to a World's Fair.

You have written, "The drive to make a machine-person" is irresistible; you say it's the "culminating tour de force of the history of technology and the history of art, simultaneously."

That's true, and it's the motivating force behind artificial intelligence. Artificial intelligence has already come up with a lot of powerful and valuable work and will come up with a lot more. But there's a difference between saying that and saying ultimately software and culture will run together. Physical stuff is too important.

Will computers be able to emulate and recognize emotions?

It is inevitable. Any real emulation of human thought will have to include emulation of human emotions. In practical terms, there are many reasons why it's good to fake emotions in software. However, there's a difference between an emulation and the real thing.

Will a computer ever be able to recognize beauty?

It will be able to identify but not recognize. You can build a computer capable of assigning a number on the beauty scale that corresponds roughly to what a human might do. However, I don't think we will ever come up with a computer that cares.

"[The Mac desktop] is beautiful because the technologists who built it aimed to make it beautiful, and thought of themselves as artists."





"Microsoft got famous selling serious, ugly software, outfitting the 'man's computer designed by men for men.'"

You tend to talk about beauty as if it were a given, a universal. Couldn't the parallel be language? All humans come equipped with linguistic ability, but we speak a thousand different languages, all of which evolve over time. Why not the same for beauty?

It's a matter of dealing with the facts on the ground, which I think are astonishing. You can say the sense of beauty evolves and reflects social constructs, but people read and love Homer today. For them to love a poem that originates in the Middle Bronze Age, or the great writing of the Bible, or classical Athenian art -- products of civilizations dizzyingly different from ours -- is remarkable, and demands an explanation.

When Europeans went to Africa in the nineteenth century they heard African percussion as either devil's work or nonsense. Something changed in the Western mind for us to appreciate its beauty. Likewise, for African sculpture.

A hundred years ago there was a huge fad in Paris for Japanese art. It is always a characteristic of Western culture to be fascinated by other cultures, never a characteristic of other cultures. There have been fads for Japanese art, Persian art, Turkish art, African art. All these have had their day in the sun. It's a characteristic of modern intellectual life to denigrate Western culture and be especially interested in non-Western culture.

Couldn't we say that we have access to many cultures today, and that that's got nothing to do with self-hatred or denigration of Western culture?

"We ought to start teaching Velázquez, Degas, and Matisse to young technologists right now on an emergency basis."

The equal-basis hypothesis is dogma that many people reject. The claim that the greatest African art is as good as Michelangelo is something I think is false. I don't think that there is any African art that reaches the heights of Beethoven. As far as I go, there is nothing outside Western art that can compete with the greatest and deepest Western art, which is why you see Western art triumphant all over the world, why the Tokyo Symphony and the Tokyo Quartet are great Beethoven specialists, why shows of Western art in the Orient are increasingly well-attended, and why Japanese and Taiwanese collectors pay thirty-five million dollars for Van Gogh. There is a hierarchy of values. It's easier to deny it, but that doesn't mean it doesn't exist.

Let me get at the question of beauty from another angle, by saying that quantum mechanics is not beautiful in any traditional way.

As a body of teaching, it's arguably not, and, of course, many physicists have been bothered precisely by its complexity. But there's general agreement that even though quantum mechanics is correct as far as it goes, it's not complete. There are more theories to discover. It's the very complexity and ugliness of the existing theory that has people bound and determined to go beyond it. It's generally the case that ugly theories with pragmatic utility are replaced in the long run by more general and more beautiful ones.

You believe moral relativism these days is a result of the intelligentsia taking over the media and dissolving traditional values.

If I look at the press in 1940, it's radically different from today's. If I look at intellectuals, they're the same. Intellectuals in 1940 said all the things everybody says today: tolerance is the single trump-all value; any family structure is as good as any other; libertinism is the way to go; sex roles should be interchangeable. Intellectuals have been saying this for decades. The difference is now everybody says it. NBC says it, too.

Isn't it a problem that what you call traditional values were being upheld by an elite, a very closed elite? In what sense is that a useful politics to uphold today?

We can't go back to 1940. It's not an option. But in 1940 there weren't major cities where murderers were never prosecuted. In 1940 public schools weren't failing and held in contempt. It's just a matter of historical fact that this country has slid downhill disastrously in many ways.

You're talking about an age when Jews were excluded from many colleges, and Jim Crow was still in force.

That goes without saying.

But it needs to be said. Maybe we're in a place where a new consensus has to be arrived at now that so many more are included.

We are in a transitional time, I agree. It certainly used to be the case that the country was run by a WASP aristocracy and no longer is. Now, when the WASP aristocracy ran the country, in certain respects the country was worse off, and those respects are undeniable -- and we hear about them ad nauseam, so we can never forget them. We know about racism, we know about anti-Semitism, and nobody denies it.

Here's what some people do deny. Some people do deny that when the WASP aristocracy ran the country it was in some ways better off. It's crucial for me to know the country was in some ways better because I can't allow my children to grow up thinking this is the best it gets. They have to know street crime is unnecessary. They have to know that divorce rates of the sort we have today are unnecessary. They have to know that public schools that are a joke aren't necessary and didn't used to exist. A lot of people today, particularly liberals, don't know that and don't believe it.

You focus on the press. But mass media today is electronic. And isn't it the nature of electronic media to undermine traditional values?

Why shouldn't newsreels in 1940 have had the same corrosive effect?

It took a while for it to happen. I don't think the intelligentsia played much of a role at all; media eats away at all restraints.

There is a historical conundrum that needs explanation. I don't care how you explain it. You may think my explanation is wrong.

I do.

It's an astonishing thing that we should have gotten so much richer, so much more powerful, so much more in command of technology -- and yet have screwed up in so many profoundly important ways. How could we have done this? How could the country make such material progress and yet wind up with a society dominated by failing schools and failing cities and disintegrating families?

You may think my explanation is wrong. Fine. But we have an obligation to come up with an explanation for this amazing historical fact. Vast numbers of people, particularly my colleagues at universities, won't even ask the question.

How does Machine Beauty address this issue?

I don't think it does. It's important that issues that are ultimately political not be allowed to take over everything. I think it's important that you not have to be a right-winger to want to read something I wrote. It would be a catastrophe if that were true of me or of anybody.

"To hate technology is in the end to hate humanity, to hate yourself...."

One question about the Unabomber.

Go ahead.

What do you think it was about your work that drew him to you?

I don't have any idea. I couldn't care less. I refuse on principle to speculate about the man's motives.


See a series of excerpts from David Gelernter's Machine Beauty: Elegance and the Heart of Technology.

Join the discussion in the Digital Culture forum of Post & Riposte.

More on Digital Culture in Atlantic Unbound

At a critical point in his life, Harvey Blume chose English over C and therefore writes reviews, criticism, and even the occasional book rather than computer programs. The co-author of
Ota Benga: The Pygmy in the Zoo (1992), he writes about art, literature, and new media.

Copyright © 1998 by The Atlantic Monthly Company. All rights reserved.
Cover Atlantic Unbound The Atlantic Monthly Post & Riposte Atlantic Store Search