Selected reviews of books about science, 1992-2000
Reviews by Damien Broderick

VIRTUAL WORLDS: A Journey in Hype and Hyperreality

by Benjamin Woolley , Penguin, 274pp

MIRROR WORLDS

by David Gelernter, Oxford University Press, 237pp

THE METAPHYSICS OF VIRTUAL REALITY

by Michael Heim, Oxford University Press, 175pp

THE VIRTUAL COMMUNITY: Finding Connection in a Computerized World

by Howard Rheingold, Secker & Warburg, 325pp

When William Gibson visited Australia recently, the coiner of `cyberspace' revealed that he isn't a denizen of that virtual realm. He doesn't even have an e-mail address--due, he remarked wryly, to `cyber-agoraphobia', a fear of being pressed to death by a global flood of incoming electronic messages. Not that cyberspace is new. Gibson's sometime collaborator Bruce Sterling, in a pioneer book on the topic--The Hacker Crackdown, now in Penguin--surprisingly defined it as `the "place" where a telephone conversation appears to occur'.

            But the true home of virtual reality (VR) is the interactive space `inside' specialised computer programs where one or many humans are immersed in a hallucinatory alternative world. We may touch, see, reshape a vivid electronic universe--indeed, fly through it, teleport instantly between virtual locations, or drastically transform the way others see us. As desk-top computers and modems proliferate in offices and homes, this virtual real estate may soon change our lives drastically. Unlike Bill Gibson, we might not have the luxury of staying out of cyberspace once its portal is there on every desk.

            Sceptics shrug off such apocalyptic visions. British journalist Benjamin Woolley pulled on datagloves and eyephones and was bitterly disappointed. `I had experienced a crudely rendered, primary coloured series of badly coordinated images. I got none of the promised sensation of liberation... just frustration at the unresponsiveness of the equipment.'

            These are, though, distinctly early days. It's almost impossible to get a gut-level grasp on the speed of change in cyberspace. Rheingold, a cyber enthusiast and author of 1991's excellent Virtual Reality, notes one index: for more than a decade, the US military network ARPANET linked its users at 56,000 bits per second. Call that 1000 words of text. (An early PC modem mumbled at 300 bits per second.) In 1987, a successor network sped to a million and a half bits per second, and by 1992 that had jumped to 45 million bits through a given line. A leap from around four typed pages to five or six hefty books per second! Systems under test run around a billion bits per second; trillion-bit lines are expected.

            Such numbers are paralysing. We slough them off. But look: it's as if the speed of jetliners had zoomed from around 500 kph to 400,000 kph, with prospects of domestic flights at the speed of light. Woolley's sub-title is right--we're poised somewhere between advertising hype and genuine hyper-reality, the condition of the postmodern. Luckily, there's now an abundance of books (quaint old fashioned things) about the opportunities and threats of virtual worlds.

            If Woolley's amusing survey is an excellent primer--ranging with fair accuracy from computer simulation and artificial intelligence to hypertext, postmod fiction, and back to (quantum) reality--David Gelernter's is the real thing made easy to swallow. A major researcher at Yale, Gelernter is striving passionately to build computerised mirrors of our human world, cyberspace models we can inspect at varying levels of detail and immersion. Chips are cheap, and getting cheaper. So are interface media such as cameras. Our public world could be monitored automatically (with due regard for personal privacy... hmm), and the torrent of data sorted into a kind of Jorge Luis Borges map, a `public infomarket', intricate as reality and open to everyone's probing curiosity.

            The joy of Gelernter's book, almost inevitably reminding one of Douglas Hofstadter's playful essays, is his blend of silliness and rigour. Even if you've never thought about smart machines before, even if you've rarely touched a computer keyboard, you'll enjoy this excursion into the Tuplesphere (or, in Michael Heim's borrowing from Heidegger, the Gestell or information framework). A `tuple' is a packet of information adrift on a cyberspace Gestell crawling with hungry infomachines, software agents hunting for requested items of data and bunching them together.

            In the Tuplesphere, data trees grow like nerves on Trellises as a rush of input at the lowest levels gets sorted, abstracted and sent upward toward command levels where machines and human intermingle. It's here that `topsight' is gained, a grand view of the structure we've chosen to inspect--local government or police in action, say, or the ecology of a river, or the history of red shoes--forests previously hidden by trees. Functioning in this new kind of space, navigating Gelernter's `recursive free-form dollhouses' of data--is done with code devices kids'll soon learn along with the alphabet: `plunge' and `squish'.

            When we plunge into the Gestell or memory pool, we ask the system to grab all the tuples (or data objects) with certain specified characteristics. They come in plenty of shapes and sizes, of course, and need a good squish to squeeze out irrelevancies, leaving behind a batch of items united into the desired category. Because a long-suffering machine is doing the plunging and squishing, new aspects of the world can emerge through these simple procedures, which to some extent evade human prejudice. Such machines already assist medical diagnosis (and are often more accurate than humans alone).

            Will this new life-world of organised information change us as human beings? Michael Heim declares himself a `techno-Taoist', bringing Heidegger's arcane philosophy of Being to bear on an emerging metaphysics of machine-mediated thought. While he raises all the obvious issues, and a few unexpected ones (linking cyberspace to Leibniz's ambition to create a universal logical calculus), he never quite gets to his goal. Still, in an admiring introduction, VR pioneer Myron Krueger praises Heim for recognising that virtual reality is `the first intellectual technology that permits the active use of the body in the search for knowledge'. He's right: despite the fearful cries of people alarmed by shambling, body-hating hackers and computer games where the mind seems gleefully disconnected from the `meat' body, powerful VR will enhance the body. Instead of sitting numbly before a keyboard, we'll use limbs and senses--machine-linked, it's true--to cavort in cognitive space. Oddly, Heim might not agree, maintaining that `a virtual world needs to be not-quite-real or it will lessen the pull on imagination'.

            Far short of full immersion in VR is the current experience of the virtual communards of Rheingold's new book. Hundreds of thousands of people around the planet are already linked into what Al Gore calls the `information superhighway', except that the proposed American (and Australian) corporate version sounds more like a super mall for sedated couch shoppers. Rheingold is a '60s hippie utopian with a salutary sense of the risks in this newest technology. Surfing the Internet is compulsive and empowering for some, not all of them by any means nerdish hackers. As bulletin boards and special-interest networks multiply, so do the unexpected and sometimes anarchic uses of cyberspace.

            The French government's Minitel service (millions of home terminals replacing costly phone books) was swiftly hacked and opened for party-line on-screen conversation among subscribers. Instantly, the notorious messageries roses erupted, a computer equivalent of feverish phone sex, and nearly crashed the system. Many of the world's academic and other nets have been colonised by MUDS (Multi-User Dungeons, role-playing games in virtual landscapes and banned in Australia) replete with `net.sleaze', `tinysex' and other boyish fantasies.

            If mature cyberspace opens the way to a new kind of democracy based on frequent electronic plebiscite (dubbed `the Zeitgeist Machine' in 1966 by Australian alternative publisher Phillip Frazer), it is also a conduit to `disinformocracy'. CMC (computer-mediated communications) and its communities revealed by Rheingold are intoxicating and liberative, but they run every risk of being channelled into controlled commodities, taxed out of existence, or perverted into a Panopticon surveillance system that'll make 1984 look like 1948. For now, the Gestell is growing rapidly in virtual space. We have almost reached the day, as Gelernter's sub-title puts it, when `software puts the universe in a shoebox'. It's up to us to learn how to dance rather than shop or blob out in its digital terrain.

SIGNS OF LIFE: The Language and Meanings of DNA

by Robert Pollack, Viking, 212pp

DID DARWIN GET IT RIGHT? Games, Sex and Evolution

by John Maynard Smith, Penguin, 264pp

THE NEW SEXUAL REVOLUTION

by Robert Pool, Hodder Headline, 310pp

NATURE'S MIND

by Michael S. Gazzaniga, Penguin, 220pp

ARTIFICIAL LIFE: The Quest for a New Creation

by Steven Levy, Penguin, 390pp

The late Paul de Man, saint of American deconstruction, proposed that `nothing, whether deed, word, thought, or text, ever happens in relation, positive or negative, to anything that proceeds, follows or exists elsewhere, but only as a random event'. Perhaps this mad idea is less breathtaking in view of de Man's youthful pro-Nazi writings, which he never acknowledged. But it also underpins a widespread critical claim: that language is radically indeterminate, so that no text can validly be given an authoritative reading.

            It's often said that the DNA of our cells comprises a message, written in a `genetic code' or `language of the genes'. Via a dizzying chemical virtuosity, its four-letter alphabet and three-letter words construct our tissues. Each of us is `written' into existence, within a specific, rich cultural environment, from a single recipe of three billion letters, or 100,000 genes. The Human Genome Project will have transliterated a standard version of this document by 2006.

            So if we are all DNA texts, might molecular biologists and neuroscientists be best advised to move their labs to the English Department? Robert Pollack, a specialist in the SV40 virus, says science has much to learn from current literary theory. It's unlikely that he'd go as far as de Man in destabilising the text. While mutations and other forms of scrambled code are crucial to evolution, even more necessary is the strict specificity of each folded protein. Errors in copying or expression are often lethal.

            Unlike human language, where our words are largely arbitrary (a sheep is a mouton is an oavtsa), `a protein is the meaning of a DNA word'. Individual genes usually come in a handy range of options, or alleles, making even close relatives an unpredictable mess of mix-and-match characteristics. What's more, the same gene cluster in a spleen cell will do a different job if a brain cell activates its message--an ambiguity that Pollack compares with literary interpretation.

            Some readers will find this act of intellectual hybridisation dazzling, and it's true that Pollack writes compellingly on an opaque topic. I found his parallels forced, though I approve of his final message: that science, no less than literature, needs to approach its DNA texts as documents with a history. It is a message de Man's method tends to erase.

            Science depends upon another kind of language, one just as unforgiving as the protein code. Maynard Smith captures this feature amusingly. The strangest feature, he notes of two books by selfish gene evolutionist Richard Dawkins, is `that neither book contains a single line of mathematics, and yet I have no difficulty in following them'. 

            Maynard Smith writes essays less indulgent than, say, Stephen Jay Gould's, sharing a crisp exactness with the late Sir Peter Medawar. Unforgivably, his publishers do not make it clear that most chapters are reviews of notable recent biology books--by Dawkins, sociobiologist E. O. Wilson, his foe Philip Kitcher, Nobelists Ilya Prigogine and Manfred Eigen, a score of others. These titles are listed at the end, but the chapters don't make much sense until you tumble to this unannounced fact. It's a shame, because browsing Maynard Smith's essays is like eavesdropping on brilliant tabletalk, his musings on the most intriguing evolutionary theorists of the last decades. He aptly captures the `essential features of good science writing: it tells me of facts and ideas that are new to me, and it makes me want to argue with the author'. And of troubling topics such as sociobiology, he notes wisely: `I tend to find myself disagreeing most strongly with whichever side I talked to last'.

            That surely can be said of those who discuss the politically impossible topic of inherited cognitive and personality disparities between the human sexes. Still, the material Robert Pool marshals is crucial reading to anyone hoping to deal with the intersection of social pressures (including language) and our genetic makeup. Infuriatingly, his detailed book has no index.

            Soon, in virtual reality, people will relate to each other convincingly in personae of the opposite sex (or indeed of animals or intelligent mists). Will optional presentation of self finally obliterate the socially constructed gulf between males and females? Pool's evidence suggests powerfully that it will not. The language of the genes is inflected by hormones, especially those implicated in reproductive dimorphism--whether a child is to be a man or a woman, or some mix of the two. The evidence from subtle brain scans and long term sociological studies argues that, statistically, male and female brain structures take alternative developmental pathways. Thereafter, in Camilla Benbow's summary: `Women are drawn to people-oriented fields, men toward objects'. Too dangerous to think about? `Women have been monstrously misused,' as Maynard Smith notes, but `the cause of women will not be helped by refusing to think dispassionately about the nature of sex differences'.

            An extreme example: famously, an identical twin boy was accidentally mutilated at seven months during circumcision. Snipped and tucked, she became a (rather tom-boyish and pushy) girl. For many years her case confirmed the priority of culture over nature. Well, no. In 1993, it was announced that in adolescence she'd chosen to have plastic surgery `to rebuild a scrotum and penis'. His brain, it's surmised, had been too `masculinised' in the womb for her to live happily as a female adult.

             This new material--most of it collected meticulously by feminist women neuroscientists, not mad-dog patriarchs--calls for very careful study. Some women reject it out of hand as `backlash' persiflage. If Pool's claims are correct, the consequences will be sizeable. Horrifying though I find it, as a greybeard old '60s anarchist, it might turn out that (in some respects at least) men and women are indeed complementary rather than equivalent. Pool, avowedly even-handed and liberal, lets his own guard down in a revealing slip. The female spotted hyena, he remarks, `sports a penis. Well, it's not literally a penis--it's merely a clitoris so enlarged that it looks like a penis'. Merely a clitoris...

            This renewed interest in, as they say, `bottom up' accounts of human functioning is supported by two other mesmerising lines of research. Michael Gazzaniga is the neuroscientist whose work on `split brain' patients startled us two decades back, showing that right and left cerebral hemispheres have quite different talents. We blend, it seemed, two minds: one verbal and logical, the other spatial and intuitive. This, of course, mixed the ingredients commonly seen as male versus female.

            Gazzaniga's continuing work shows that men and women are specialised in their brain architecture. Far more shocking, I suspect, is his claim that what we can know and do is largely pre-set by the brain's wiring. His analogy of selection rather than instruction (also advanced by Nobel laureate Gerald Edelman) proposes that mind resembles the immune system. From infancy we contain billions of variant antibody molecules, each primed for its antigen of choice. Just so, our brain is built from specialised modules (rather like `faculties'), interacting through an `interpreter' region. Linguists accept this model for grammar; if Gazzaniga is right, it applies as well, as his subtitle puts it, to `the biological roots of thinking, emotion, sexuality, and intelligence'.

            The ultimate outcome of such investigations is the quest for life-forms simulated on a computer. Stephen Levy's readable and wide-ranging survey makes this newest scientific project accessible to everyone. Like `complexity theory', an adjacent discipline, artificial life hints strongly that neither `bottom up' nor `top down' methods--neither the reductive nor the holistic--provides the whole truth. Self-replicating strings of numbers, or genetic algorithms, can be set free to evolve in the virtual ecology of a machine. By sheer brute Darwinian power, their lineages (like the `data-farms' predicted 15 years ago by Rory Barnes and me in our novel Valencies) yield powerful new programs designed by no human mind.

            Are these wonderful routines `alive'? By some standards they have more right to the word than viruses do. They are the very contrary of de Man's beliefs about language: arbitrary yet lawful, coherent, combative, striving to declare their selfhood in a language of simulated DNA that, as this fecund century draws to a close, speaks to us all too clearly.