The Science of Avatar

by Stephen Baxter

  The vast numerical superiority of the Inca meant nothing in the face of the Spaniards’ technological advantage. The Spaniards were a gunpowder culture facing essentially a stone age civilisation. Their steel weapons slashed through the thin armour of the Inca. And the Spaniards’ use of hourses terrified their enemy. As the horse had long been extinct in the Americas, when faced with cavalry charges the Inca did not even understand what they were seeing (remember Captain Cook’s ship and the islanders). Worst of all, in subsequent decades the “herd diseases” like smallpox that the Europeans inadvertently imported from home caused a huge implosion of the native populations.

  This basic pattern, of the overwhelming advantage afforded by superior technology, and the leveraging of that advantage into conquest and exploitation, appears to be a common theme of human history. It goes on today. James Cameron intended Avatar as, in part, a cautionary tale about the consequences of contact, colonialism and exploitation. Cameron and some of the cast of Avatar visited the Xingu people of Brazil, who live in a part of Amazonia likely to be affected by the multi-billion-dollar Belo Monte hydroelectric dam project. Cameron calls this a “real-life Avatar confrontation… in progress.”

  In the past it has even happened to “us” in the western world. Britain was overwhelmed when the Romans arrived, with their superior army discipline, road-building and literacy-based communications. For all the supposed advantages of Roman civilisation that followed—and Britain’s subsequent history is unimaginable without the Roman intervention—it wasn’t a comfortable process to live through, as Queen Boudicca (Boadicea) of the Iceni nation demonstrated in her bloody but futile revolt a generation after the Romans landed.

  If it happened to us before, could it happen again in the future? By the end of the nineteenth century one thoughtful witness, H. G. Wells, disturbed by the plight of peoples like the Tasmanians who appeared to have been entirely exterminated during European colonisation, wondered how it would be if humans, specifically the Victorian-era imperial British, were ever on the receiving end. In The War of the Worlds, British army guns facing the Martian heat ray are “bows and arrows against the lightning”—a phrase evocative of the battle scenes of Avatar.

  Today some like to imagine, as in Carl Sagan’s Contact, that if the aliens come we will receive wisdom from the stars: an Encyclopaedia Galactica, a cultural adrenaline boost that will raise our society to new levels. But others follow Wells in imagining harsher possibilities. Physicist Stephen Hawking recently said (in a Discovery Channel documentary called Stephen Hawking’s Universe, aired on 9 May 2010), “I imagine they exist in massive ships, having used up all the resources of their home planet. If aliens ever visit us, the outcome could be much as when Christopher Columbus first landed in America, which didn’t turn out very well for the Native Americans.” Which sounds like an Avatar scenario in reverse.

  Today there is a ferocious debate going on in the world of SETI, the search for extraterrestrial intelligence, about the wisdom, not just of passively listening for signals from space aliens, but of signalling to them. This is known as “active SETI.” Earth is a noisy place in the radio spectrum; we’ve been leaking radio, TV and radar signals for decades. But the signal strength drops off quite quickly, over a few light years, spanning a few tens of stars, say. Purposeful signals would suddenly make us visible to a much larger chunk of the Galaxy. And signals have been sent before. In 1974, the Arecibo radio telescope in Puerto Rico transmitted a series of radio pulses towards the M13 star cluster, encoding a message from humanity designed by SETI pioneer Frank Drake.

  Some have always been unhappy about this. Former Astronomer Royal Sir Martin Ryle warned that “any creatures out there [might be] malevolent or hungry.” And Sir Bernard Lovell, founder of Jodrell Bank, once said, “It’s an assumption that they will be friendly—a dangerous assumption.” Science-fiction writer David Brin speaks of analogies of toddlers shouting in the jungle. Maybe this is the resolution to the Fermi Paradox: everybody else keeps quiet because they know there is something dangerous out there.

  But does it have to be this way? Is it in us to learn to love the alien? And could the alien ever love us?

  At least we know we ought to behave better.

  The “Golden Rule” of ethics, which is embedded in many religions and philosophies, was expressed by Christ this way: “Do unto others as you would have them do unto you.” (This wording, a version of verses from the gospels of Matthew and Luke, first appeared in a catechism in the sixteenth century.) Also known as the “ethic of reciprocity,” the Golden Rule is arguably the basis for the modern concept of human rights: that you should treat everybody, including those not in your own immediate allegiance group, with consideration. It has been criticised. George Bernard Shaw pointed out that the other’s taste may not be the same as yours; how do you know that the others would like having done unto them what you want done unto you. But nevertheless it’s not a bad principle to live by. As Wells pointed out, the imperial British wouldn’t have enjoyed having the Martians doing unto them what the British did to the Tasmanians.

  Even during the darkest years of the European colonisation age, there were flickers of empathy. As early as Columbus’ own expeditions, some people back home were appalled by accounts of slavery and massacre. It wasn’t long before the Pope decreed that the Native Americans were fully human, that they had souls, and that the mission of Christians must be to save those souls rather than exploit their bodies. The Christians missionaries that followed did a good deal to disrupt and destroy native culture, but in the context of the sixteenth century I think you must call the Pope’s decree a hopeful sign.

  Interesting debates continue today, incidentally, about the theological status of hypothetical extraterrestrial aliens. There is no sign of any Christian or other missionaries working among the Na’vi. The collision with Eywa would be fascinating. Perhaps it might help the Na’vi’s cause if some twenty-second century Pope in faraway Rome were to declare that they too have souls…

  But the Na’vi aren’t exotic humans, like the Native Americans. They are alien creatures. We can empathise with human strangers; could we ever empathise with the alien?

  Again, precedents from our career on Earth aren’t very hopeful. Consider how we treat the animals. Though the Na’vi respect the animals they take for food, on Earth even our closest surviving relatives, the great apes, are in danger of being driven to extinction through the carelessness of habitat loss and fragmentation—and, sadly, from purposeful hunting.

  We tend to measure animals’ worth in terms of how much they are “like” us. Thus we look for signs of human-like cognition in chimps, as expressed in tool-making and sign language. But maybe, as philosopher Jeremy Bentham said as long ago as 1789, we should treat an animal depending not on how well it thinks but on how much it is capable of suffering. Consider the heartbreak of a mother elephant when her baby is taken by the poachers. Scottish psychologist James Anderson has compiled data on how chimps treat their dead. Mothers can carry corpses of their dead babies around for weeks, even though it is clear from subtle reactions that they know the infants are dead. Such observations “make a strong case that chimps not only understand the concept of death but also have ways of coping with it,” Anderson says.

  You only have to consider your own feelings when watching the distressing scenes that follow the aftermath of the destruction of the Na’vi’s Hometree to believe, I think, that we will indeed be able to empathise with the alien, when we meet it. After all, though in Avatar there is a Miles Quaritch, there is also a Grace Augustine, wanting to reach out to the Na’vi.

  But it would surely make it easier to empathise with an animal if you could plug your mind directly into it.

  28

  MIND TO MIND

  One way in which the Na’vi are entirely unlike us is in their queues.

  A queue is a hair braid encasing a neural whip, an intricate mass of active neural tendrils. The Na’vi are able to join thi
s organ to similar structures on other animals to make a neural bond, which the Omaticaya call the shahaylu. Onscreen we see this work with the direhorse, the banshee and the leonopteryx. Through his bond with his direhorse, avatar-Jake can sense the animal’s body, her heartbeat, her breath, the strength in her legs. And his will to some extent overpowers that of the linked animal. At first he commands her with words, but ultimately he is able to control the direhorse with inner “commands,” just as he controls his own body.

  As well as a very visible demonstration of the Na’vi’s integration into their ecology, this is clearly a terrifically useful biological technology. It is like a natural version of the comprehensive neural interfaces that must be necessary to run Jake’s avatar body, as we’ll see in the next section. Perhaps avatar technology was in part inspired by the natural version on Pandora.

  But we might speculate that in some ways the shahaylu has stifled Na’vi cultural evolution. If you can will a direhorse into submission you don’t need to break it. Perhaps the generations-long process of domestication with which humans have filled their world with more “useful” versions of animals like horses, sheep, cattle and dogs will never occur to the Na’vi.

  What’s of more interest to us amateur xenobiologists, however, is how the neural link evolved.

  The shahaylu is even more remarkable when you consider what diverse animals it links: Na’vi, direhorse, banshee. Humans are fairly remote relations to horses, and even more remote from birds and pterosaurs. As life on Earth evolved, the family of primates that would one day include humans split off from the “laurasiatheres,” the tremendous group that includes horses (along with camels, pigs, dogs, cats, bears…) as far back as eighty-five million years ago. This wasn’t even in the time of the mammals’ dominance; this was back in the Cretaceous age, the dinosaur summer before the big impact. And we split off from the group that includes leathery flying lizards such as the pterosaurs (and indeed the birds) even further back in time: an astounding three hundred million years ago, back in the Carboniferous age, halfway back to the time of the emergence of multicellular life in the first place. If similar evolutionary gulfs separate Na’vi from direhorse and banshee, how is it possible for them to develop such an intimate link as the shahaylu?

  I can think of one terrestrial parallel: bees and flowers.

  Like a bee pollinating a flowering plant on Earth, a Pandoran direhorse has a long snout it uses use to feed on sap drawn deep from plants like the direhorse pitcher. Both halves of the partnership benefit. The direhorse gets protein from insects trapped in the sap, and the pitcher plant gets pollinated. This behaviour is shared by some terrestrial animals, such as lemurs and possums. In some senses the link between bee and flower (and direhorse and pitcher plant) is a lot more fundamental even than the shahaylu. The bee has come to rely on the plant nectar for food, and the plant entirely relies on the bee for its means of fertilisation. This cooperation is not just a temporary alliance for horse-riding, but determines life and death for both partners.

  But there had been insects around for three hundred million years before the flowering plants, the angiosperms, first appeared on the Earth back in the Cretaceous, the heyday of the dinosaurs. And the split between the plants and the vast family that includes all animals, insects and fungi was extraordinarily far back in time—billions of years ago. But once the flowering plants emerged, they co-evolved with the insects they cooperated with, establishing their extraordinarily intricate interdependence over millions of years.

  So it is possible for astonishingly distantly related species to develop a remarkably close degree of cooperation, given enough time for natural selection to work. Perhaps something like this lies behind the shahaylu.

  But for a Na’vi, whatever its origin, the neural queue is intimately linked to her experience of sex and death.

  To quote the 2007 screenplay: “Neytiri takes the end of her queue and raises it. Jake does the same, with trembling anticipation. The tendrils at the ends move with a life of their own, straining to be joined… The tendrils intertwine with gentle undulations. Jake rocks with the direct contact between his nervous system and hers. The ultimate intimacy. They come together into a kiss and sink down on the bed of moss, and ripples of light spread out around them…”

  The love-making between avatar-Jake and Neytiri is the culmination of their strange courtship. The joining of their neural queues is fundamentally involved; this is a joining of minds, of consciousnesses, as well as bodies. And as we see onscreen the outcome of this joining is a lifelong, irrevocable bond, cementing the culture’s monogamy.

  But for a Na’vi warrior a queue is also a weakness. At the climax of the battle between RDA and the Na’vi there is a brutal fate for the warrior Tsu’tey, when a human soldier brutally cuts off his queue. The human has heard this is “worse than death” for a Na’vi. Perhaps it would be. Tsu’tey could no longer ride a direhorse or banshee. He would be excluded from sexual intimacy; this is a symbolic castration.

  And, worse, Tsu’tey will suffer a deeper death than his ancestors. For his queue is also his connection to Pandora’s greatest mystery of all: Eywa. And through her, immortality.

  29

  EYWA

  As Selfridge and Quaritch prepare to use lethal force against the Omaticaya and their forest, Grace Augustine protests, trying vividly to express what she believes she has learned of Eywa.

  Grace has found evidence of “electrochemical communication” between the roots of Pandora’s trees, similar, she thinks, to the synaptic sparking between the neurons in a human brain. This is the basis of a natural neural network, like a human brain, but on a planetary scale: Eywa.

  To the Na’vi, Eywa is their mother goddess—and, in a sense, their heaven. She takes them into herself when they die. We see this when the transfer of the essence of dying Grace to her avatar body is attempted in the Tree of Souls. This fails—but “all that [Grace] is” is taken into Eywa. Some essence of the dead Na’vi survive within Eywa, and the living can communicate with them by plugging a queue into a natural “portal” such as the Tree of Voices. This is why the amputation of Tsu’tey’s queue was so cruel; worse even than murder, it denies him immortality among his ancestors.

  If Eywa were a human computer, all this is plausible if you believe in the much-anticipated technologies of “mind uploading”—mapping the brain and transferring its contents to a computer store—which we will look at in the next section when we investigate the avatar link process itself. But Eywa isn’t a Cray supercomputer. She has no silicon chips or optical links. She’s not even a human brain, a mesh of intricate biochemistry. As Parker Selfridge protests, “What the hell have you people been smoking out there? They’re just. Goddamn. Trees.”

  Is it really plausible that a bunch of trees could be connected up into a network with anything like the “functionality” of a brain? And even if you believe a forest can become a brain, how smart can it possibly be?

  Let’s start with Grace’s numbers.

  Grace says that each tree on Pandora has “ten to the fourth” connections to the trees around it, and there are “ten to the twelfth” trees on the moon. This, she says, adds up to a global neural network with more connections than the human brain.

  Can there really be “ten to the twelfth” trees on Pandora? Ten to the twelfth power means ten multiplied by itself twelve times, a number you’d write down as one followed by twelve zeros, with commas scattered according to taste: 1,000,000,000,000. That’s a million million—a trillion. By comparison, how many trees are there on Earth? In 2008 Nalini Nadkarni of Evergreen State College in Washington published an estimate, based on NASA orbital images of forest coverage. Her number was absurdly precise: 400,246,300,201—four hundred billion, or around sixty trees for every person on the planet. That’s about half of Grace’s estimate for Pandora. And given that much of Earth has been deforested by us humans in the last few thousand years, that’s close enough for me to accept Grace’s number as pl
ausible, even though Pandora is smaller than Earth. Meanwhile the human brain is believed to contain around a hundred billion neurons—that’s ten to the power of eleven in Grace-speak. That’s a factor of ten less than Grace claims for Pandora’s tree number. As regards connections, on average a brain neuron has about a thousand connections to neighbouring neurons: that’s ten to the third, again a factor of ten less than Grace claims for the trees.

  So your brain amounts to a total network of around a hundred trillion connections (a hundred billion times a thousand). And on that count, Grace is right that the network on Pandora is indeed bigger than the human brain, by a factor of about a hundred.

  How does this compare to modern computers? Each neural connection in your brain can support about two hundred “calculations” per second. So that’s a total of twenty thousand trillion calculations per second, going on in your head, right now. (Granted it may not always feel like it.) That is, in the information technology terms I used in Chapter 19, the brain is capable of twenty petaflops. As we saw in Chapter 19, as of 2010 the most powerful non-distributed computer system in the world, the Chinese “Milky Way,” was capable of 2.5 petaflops—an eighth of the processing speed of the human brain, or about a thousandth the power of Eywa.

  That sounds impressive, but you’ll recall from Moore’s Law (Chapter 19) that computing speeds and capacities are multiplying rapidly, doubling every fourteen months according to the TOP500 study. If this goes on, the fastest computer will pass the brain for sheer speed in just four more years—and it will pass the more powerful Eywa in a mere dozen years.

  Or will it?

  Grace’s estimate of Eywa’s complexity is based on a count just of the physical neural network of trees and their connections. This is quite reasonable from a scientist’s point of view, since it is all that Grace can “sample” and measure. But there is more to the complexity of any computer than a simple count of the links between its components.