From Social Synapses to Social Ganglions:Complex Adaptive Systems in the Jurassic Age
History of the Global Brain, Part V
Howard Bloom is reflecting in this chapter why birds congegregate in huge flocks. He describes the advantages of flocks as collective learning machines and explains the main principles of these collective adaptive systems.
"How ya gonna keep 'em down on the farm, after they've seen Paree?"
For most of human history, the need to eke a living from the earth kept over90% of the human population in the countryside. But once a small numbercould produce food for multitudes, a formerly repressed desire went hog-wild - our urge to cram together. Today, more than 75% of Europeans and NorthAmericans have crowded into cities. In Belgium the figure tops 95%. This lustfor company has hit the developing world even harder. In a measly twogenerations, Mexico's urban congregants have leaped from 25% to 70% of thepopulation. Mexico City is now jammed with 27 million human beings,roughly three times the worldwide number of Hominids alive at even thelushest moment of the Paleolithic age.
Many species of birds are as attracted to their equivalent of the big city as weare, and given the chance, will congregate in the largest clusters they canpossibly form. Some bird flocks outdo the largest human municipalities bya factor of two - reaching 50 million or more. This sociable overcrowdingseems to court extraordinary risk. The larger the flock, the larger theterritory it must cover to feed itself, and the greater the chances ofencountering a famine. So why do avians become hypnotized by the urge tojoin a crowd?
The first guess ornithologists came up with was warmth. In winter, theyreasoned, the birds could huddle, providing each other with protection fromfreezing cold. When researchers compared the energy costs of joining a roostto the energy saved by communal heat, the results were rather surprising. Ifthe roost is thickly populated, the daily distance from home base to food islikely to involve an arduous commute. The calories burned in travel by faroutweigh the pittance saved by toasty snuggles, swallowing 27% of a starling'sentire intake for the day. Overnighting alone in a sheltered hollow - despitethe need to generate extra body heat - would exact nowhere near that price.
Why then, do birds congregate in avian megalopolises? There is something farmore critical than energy to be gained - information. Birds rely for theirperception of the world on those around them. If you recall the experimenton imitative learning among octopi from our previous episode, this will soundlike deja vu all over again. Experimenters put a young, inexperienced blackbirdand an older, wiser flier in cages side by side. The savvy elder was shown anowl, and attacked the potential killer furiously. The youngster couldn't see thepredator. Sly experimenters had placed a partition in his line of sight. But hedefinitely could witness the emergency response.
Not that there was nothingto surprise the junior bird as well. On his side of the opaque divider appeareda stuffed honey eater, a congenial creature which does not feast on blackbirdmeat. The setup was designed to convey the impression that the elder'spugnacity had been roused by the harmless sweet-snacker. Later the youngbird was put next to an unseasoned fledgling like itself. Both were shown thehoney eater. The newcomer was indifferent. But the bird who'd seen his eldergo into a rage flew at the bee-juice connoisseur, assaulting it with might andmain. Soon the novice picked up the message and joined in. Then it, too, waspaired with a naive bird who couldn't have cared less. Like his teacher beforehim, the bird who'd learned his lesson demonstrated the importance ofmobbing honey eaters to his pupil, passing the tradition on. Erroneous as itwas, this response was reproduced in six blackbird generations before theresearchers called it quits.
OK, birds have imitative learning. What's so astonishing about that? We'vealready shown the imitative passage of data in creatures as primitive as spinylobsters 260 million years ago. And we've explained how emulative absorptionacted like a synapse, allowing information to leap the gap from one creatureto another. But a whole new kind of information processor arises whenneurons or independent beings join more than mere bucket brigades. Huddledlike roosting birds in the brain-precursor called a ganglion, neurons can swapand compare data by the batch, arriving at something far beyond mere lineartransmission. Each adding to the mosaic, they can see the big picture. Or, toswitch from church floor imagery to that of the kitchen counter, whenkneaded, stretched and rolled by a social cluster, you never know what formsof output input will become.
In 1973, Amot Zahavi, the eminent Israeli naturalist, posited that the roostwas an "information center." From 1988-1990, John and Colleen Marzluffof the Sustainable Ecosystems Institute in Meridian, Idaho, and Bernd Heinrichfrom the University of Vermont attempted to test the notion. They focussedtheir attention on ravens (Corvus corax) living in western Maine's pine forests. Their technique was to capture wild ravens and to keep these carrionconsumers caged until all their existing knowledge about food locations wasthoroughly out of date. Then the experimenters put a fresh carcass - theultimate raven cold-cut buffet - in a previously unused site, let the birds in onits coordinates by showing them the lump of meat as the sun was setting, andset the newly enlightened ravens free. The next day, only one of the 26 birdslet in on the secret showed up - leading 30 ravens from a roost over a mileaway. During the next few days, two more of the experimentally isolatedravens also came back to feast on the cadaver. Each had a trail of roost-matesin its wake. From this and a variety of other experiments and observations,the three researchers concluded that "Raven roosts are mobile informationcentres" in which the birds, by means unknown, swap data on where succulentcadavers are to be found, then follow the bird most in the know the next daywhen the flock takes off. In addition, the ravens share their information withothers far away, engaging in a "social soaring display" which can attracthungry and clueless conspecifics from up to thirty miles.
So Zahavi had been right. Roosts, at least among ravens, are collective dataprocessors. What's more, they are part of local networks, pooling databetween strangers for the sake of all.
Somewhere between 145 million years ago, when the first feathered reptile, thearchaeopteryx, arose, and 120 mya when modern birds appeared, imitativelearning among vertebrates went from serial to parallel wiring, making a socialgroup a learning machine. The mechanism for massed learning and collectiveadaptation was apparently at work in the herding and hunting beasts we knowas dinosaurs. Paleontologist Robert Bakker hypothesizes that the herd alloweddinosaur herbivores to pool the input from their eyes, ears and nostrils, thenmount a carefully phalanxed defense. Dino-carnivores were even subtler intheir use of networking. Bakker suggests that like today's lions, they teamedup to stage elaborate stratagems. One Utahraptor might act as a decoy,distracting the attention of a brontosaurus pack. Meanwhile its hunt-mateswould surround the prey and take it from behind. But how did communallearning machines arise among Jurassic kings and queens?
To understand the global brain's anatomy as it continues to unfold, we willhave to take a side trip into theory. Specifically we've got to machete furtherdown the path of complex adaptive systems. Later we will once again resortto theory, proposing a new model of cosmic basics. But one new concept ata time.
The exploration of adaptive systems I'm offering you does NOT come fromcomplexity's Mecca, the Santa Fe institute. And unlike other theories on thesubject, it is not based on computer simulations. It is the result of 29 years offieldwork observing the real thing - social nets in action. The insights of SantaFe systems modelers like John Holland have helped me greatly in thisenterprise. But the principles I will enunciate emerge from a more elementaltechnique - that which Darwin used - venturing first-hand into the wilderness,accumulating reports from other empirical frontiersmen, and running vastquantities of data through numerous conceptual sieves in an effort to isolatenuggets of gold.
Essentials of a Collective Learning Machine
The result is a five-element dissection of a collective learning machine. Thequintet of essentials: (1) conformity enforcers; (2) diversity generators; (3)utility sorters; (4) resource shifters; and (5) intergroup tournaments.
2. Diversity generators spawn variety. Each individual represents ahypothesis in the communal mind. It is vital for the group's flexibility thatit have numerous fallback positions in the form of participants sufficientlydifferent to provide approaches which, while they may not be necessary today,could prove vital tomorrow. This can easily be seen in the operation of oneof nature's most superb learning machines, the immune system. The immunesystem contains 10(7)-10(8) different antibody types, each a separate conjectureabout the nature of a potential invader. However diversity generators takeon their most intriguing dimensions among human beings.
3. Next come the utility sorters. Utility sorters are systems which siftthrough individuals, favoring those whose contributions are most likely to beof value. These pitiless evaluators toss those who personify faulty guessworkinto biological, psychological and perceptual limbo. Some utility sorters areexternal to the individual. But a surprising number are internal. That is, theyare involuntary components of a being's physiology.
4. Fourth are the resource shifters. Successful learning machines shuntvast amounts of assets to the individuals who show a sense of control over thecurrent social and external environment. These same learning machines castindividuals whose endowments seem extraneous into a state of relativedeprivation. Christ captured the essence of the algorithm when he observed"to him who hath it shall be given; from he who hath not, even what he hathshall be taken away."
5. And bringing up the rear are intergroup tournaments, battles whichforce each collective entity, each group brain, to continually churn out freshinnovations for the sake of survival.
To understand how these five principles affect you and me, it may be helpfulto reexamine the workings of a group brain in an organism normally thoughtto have no intelligence at all: our old friend the bacterium.
Bacterial Group Brain
In the late 1980s, two scientists we've frequently met before, University of TelAviv physicist Eshel Ben-Jacob and the University of Chicago's James Shapiro,were perplexed. Those supposed lone rangers known as bacteria actually livedin colonies which established elaborate designs as they expanded. Some rippledin ringlets. Others snaked in symmetrical tracery like that generated bygraphic depictions of fractal equations.
Ben-Jacob detoured from normal physics and spent five years studying bacillussubtilis. Meanwhile Shapiro focused on such organisms as E. coli andsalmonella. Unlike the traditional biologists who had preceded him, BenJacob applied an unconventional tool to his data: the insights he had absorbedfrom the mathematics of materials science. New developments in this fieldsuggested that the elaborate patterns formed by bacterial colonies might be theresult of the same processes which produce patterning in water, crystals, soiland rocks. The Israel physicist felt that this was wrong and set out to separatethe products of "azoic" (non-living) processes from those which he suspectedwere the results of microbial hyperactivity.
Meanwhile among microbiologists another mystery was gumming up theworks. Standard neo-Darwinism said that bacteria stumble from oneinnovation to another by random mutation. But a growing body of evidencewas accumulating to indicate that bacterial mutations are not completelyrandom. Seemingly every month fresh studies continued to suggest thatthese mutations might, in fact, be genetic alterations "custom-tailored" toovercome the emergencies of the moment.
Ben Jacob confirmed what he had suspected all along. Something far morethan the principles which shape inanimate matter was at work within the petridish. Separate investigations by Shapiro and Ben Jacob uncovered a surprise,one which answered the puzzle of bacteria's seemingly purposeful alterationsand now threatens to topple long established evolutionary models. Ratherthan being a mere carrier of construction plans, the package of genes carriedby each individual bacterium functioned as a computer. What's more, thegenetic-bundle seemed to accomplish something even computers cannotachieve. Says Ben Jacob, "the genome makes calculations and changes itselfaccording to the outcome." Unlike an assembly of silicon chips, the genomeadapts to unaccustomed problems by reprogramming itself.
Reaching this conclusion left a puzzle. Godel's theorem implies that onecomputer cannot design another computer with more sophisticatedcomputational powers than its own. So how does the individual bacterium'scentral processing unit confront large-scale catastrophe, natural disaster sooverwhelming that it dwarfs the bacteria's solo computational abilities? Theanswer, Ben-Jacob hypothesized, lay in networking - in knitting the colony'smultitude of genomic personal computers into something beyond even themassively parallel distributed processor known as a supercomputer. Asupercomputer is only faster than its less sophisticated cousins, but does nottranscend many of the smaller machine's most basic limitations. At heart bothare merely diligent instruction repeaters. However the "creative net" of thebacilli, unlike a machine, can invent a new instruction set with which to beatan unfamiliar challenge.
Ben-Jacob has now analyzed thousands of colonies of bacilli to find out if hiscreative network hypothesis is true, and if so what makes the collectiveinformation-processor work. We've seem some elements of his conclusion inearlier chapters: bacilli are in constant contact, communicating through a widevariety of means, measuring their environment's limitations and opportunities,and feeding their data to each other, then finally summing the productthrough collaborative decision. In short, bacilli engage in many of the basicactivities we associate with human beings.
Here's how Ben-Jacob's work appears when filtered through the lens of asocial learning machine's five principles:
Bacillus subtilis colonies employ a variety of diversity generators. Says Ben-Jacob, bacterial clones (genetically identical offspring of the samemother) can assume intriguingly different variations. Which form each donsdepends on the chemical signals it picks up from the herd around it. Thesecues activate or deactivate individual genes, redrawing a bacteria's design andreplacing its old operations manual. In the best of times, when food isplentiful, the colony clumps together for the feast. Divergent appetites anddigestive abilities are vital to a gorging group's survival. The bacteria whichconcentrate on mining the new food source produce a poisonous by-product - bacterial excreta, the equivalent of feces and urine. Other bacteria adopt anentirely different metabolic mode. To them the excrement is caviar. Bysnacking heartily on toxic waste, they prevent the colony from killing itself.
More diversity generators kick in when the colony's glut runs out. We've already seen some of them at work in 3.5 billion year old stromatolites. As famine approaches, individuals send out a chemical signal which makesthem socially obnoxious, a "body odor" that says "spread out, flee, explore." This prods roughly 10,000 groups of cells to act as scouting parties, settingforth in a trek which unfolds before the human eye in the forms which hadfirst caught Ben Jacob's attention, concentric circles, thick fingers flaring froma central core, or a spreading circle of fractal lace. Meanwhile other cellularcohorts apparently set up posts in the wake of the outward advance andchannel the findings of the explorers toward the center.
At this stage the teams of pioneers (technically called "randomwalkers") utilize the third principle of a complex adaptive system: the colony'sutility sorters. Those exploration parties which find slim pickings have aninternal device, the bacterial equivalent of what British theorist MichaelWaller, writing about human beings, has called a "comparator mechanism." This gauge determines that the outriders have chanced across parched anddangerous territory. Their mission, in short, has failed. The unfortunatessend out the altruistic repellent which makes others in the group avoid them,leaving them to starve in isolation.
Conversely, discoverers which encounter a cornucopia of edibles havetheir comparator mechanisms tweaked in the opposite direction. Theydisperse an attractant which makes them the star of the party.
Now the fourth principle of the complex adaptive system enters thepetri dish: the resource shifters. Those stranded in the desert are deprived ofnutrients - which their location cannot provide - of companionship, and mostimportant from the point of view of the group brain, robbed of what mightbest be termed popularity. Meanwhile, those who find an overflowing buffeteat their fill and command the attention and protection of a gathering crowd. They are transformed into leaders, guiding the group mind. "To him whohath it shall be given; from he who hath not even what he hath shall be takenaway."
Should things prove truly grim, however, and even the most strenuoussearchers confirm that food is nowhere within reach, another diversitygenerator, the most startling of them all, may rouse to meet the challenge. Itis that mechanism which James Shapiro calls the "genetic engineer." Let usallow Ben-Jacob to repeat something we've already touched upon: "the cellcarries a complete set of tools for genetic self-reconstruction: plasmids, phages,transposons and too many others to mention...the same tools, in fact, used inthe lab today for genetic engineering." A microscopic research anddevelopment squadron goes to work recrafting its own genetic string.
Which raises a question: does the genomic skunkworks merely trot out pre-fabricated parts which have worked in the past? Or is it capable of trueinnovation?
This is when Ben-Jacob devised his tests of bacterial ingenuity, putting thepoor creatures into nightmare environments whose like they'd neverencountered before. If all the microbial team could do was recycle ancientprograms, it would be finished. But that is not what happened. Through datapooling, experimentation, and tests of novel strategies, the bacteria managedto refashion themselves in radically new ways. This was not traditionalrandom mutation at work. This was driven, inspired conception.
Thanks to the synergy of the conformity enforcer, the diversity generator, theutility sorter, and the resource shifter, the colony was capable of somethingnumerous humans never achieve - creativity.
In a natural environment, the fifth of a complex adaptive system'sprinciples would presumably come into play: the intergroup tournament. Alas, until recently Ben Jacob has studied each colony isolated in its own petridish, sealed off by plastic walls from competing groups. But as the resourceswhich feed the bacillus subtilis run out, imagine what might happen if a sporeof another bacterial species were to drop in, a species which found the inedibleplateau on which the subtilis was stranded to be more nourishing thansauerbraten. The race would be on. While the bacillus subtilis reworked itsgenome in an effort to gain sustenance from the now (to it) barren waste, thenewcomer would rush to reproduce, taking advantage of the fact that subtilis'inedible slabs are its entrée du jour.
As the two groups struggled to take over the petri dish, would a newinnovation emerge from the contest, an innovation of the sort which enrichesthe fate of a species for eons? One which adds abundance to the environment,complexifying the planetary biomass, transforming ever more of this oncebarren planet into food for life?
Learning Machine in Raven Colonies
We have already seen these principles at work among crayfish, birds and bees. The raven who succeeds in spotting a banquet gains followers and magnetism. It is quite likely that he also wins the privileges of hierarchical rank - first dibson mates, food, and the most comfortable overnight accommodations. Thegenes which make him a raven like his brethren are conformity enforcers. So are the tugs of imitative learning which pull him toward flying meeklywith the flock. The maverick nature which causes him to buck that impulseis a form of diversity generator. It allows him to soar over territory hisfellows have not explored, and thus to make new finds.
When his search is victorious, utility sorters shift the raven's hormonal gears,giving him internally-generated strength and confidence. Biology rewards himwith an attitude which will draw a following. Cockiness is his equivalent ofa bacteria's chemical attractor. This is equally true for innumerable species. The amount of chemotactic allure a bacteria can generate determines itsleadership. The enthusiasm of a scout bee advertising a new find determinesthe number of followers she will attract. The regal strutting of a spiny lobsterwinner almost certainly helps captivate adherents who will follow him in histrek away from a glacial freeze. Each of these creatures has been turbochargedinternally by success. And that endogenous upgrade makes all the differencein the world.
Meanwhile social machinery outside the new leader's physiological fabric setsthe resource shifters into motion, honing to unbeatable sharpness his or heredge in nutrition, reproduction and influence. Very simply put, as thechampion's hormones give him a boost, other inner chemicals downshift hisformer rivals and impel them to defer to him, funneling the group's bountyin his direction.
Finally, intergroup tournaments increase the odds that those groups whichstumble in their use of the previous four mechanisms will also fail to survive. If faulty physiology draws you to the wrong leader, you are likely to leave nogenetic or memetic legacy in your wake.
So ravens pool their findings and follow those who have demonstrated arecord of meaty discoveries and of organizational savvy beneficial to thebunch. Raven flocks even share news of their richest treasures withaggregations from miles away, as if they knew that through this worldwide-webbish generosity, they would survive the famines which permanently downthose who selfishly hog their data.
These are some of the secrets of the nascent global brain. Robert Bakker hasinferred that this quintet of principles was at work among velociraptors andastrodons 120 million years ago. New finds of early birds (Confuciusornis)from the same era also hint that the beasts with the novel feathers may haveused the five principles of a complex adaptive system in their group behavior. And we will soon see how the learning machine's pentagram extended itsembrace to human beings.http://www.heise.de/tp/artikel/2/2162/
Da bei großen Monitoren im Fullscreen-Modus die Zeilen teils unleserlich lang werden, können Sie hier die Breite auf das Minimum zurücksetzen. Die einmal gewählte Einstellung wird durch ein Cookie fortgesetzt, sofern Sie dieses akzeptieren.