Busted! The myth of technological progress
- 26 September 2012 by Laura Spinney
- Magazine issue 2884. Subscribe and save
- For similar stories, visit the Human Evolution Topic Guide
It took our ancestors 2 million years to
The history of technology holds some salutary lessons for anyone who blithely believes there is a high-tech fix for all our 21st-century problems
Editorial: "The Singularity is upon us? Not so fast"
WHEN Admiral Zheng He led his fleet out of the eastern Chinese port of Suzhou in 1405, it must have been a sight to behold. The largest of the several hundred ships under his command were the size of modern aircraft carriers and housed 500 men apiece. The fleet made seven expeditions in all, to advertise the might of the Ming dynasty around the Indian Ocean, but having returned to port for the last time it was dismantled, vanishing along with the engineering know-how that created it. For the next few centuries China's seagoing vessel of choice was a much humbler junk.
It seems incredible that such an impressive, and effective, body of knowledge could have disappeared like that, yet history is full of such examples. When archaeologists began excavating at Pompeii in the 18th century, they uncovered remains of a Roman aqueduct system that was more sophisticated than the one in use at the time. The Egyptian pyramids still haven't given up all their construction secrets. And going even further back, finds at Howieson's Poort Shelter in South Africa indicate that people were making highly sophisticated stone tools there until about 60,000 years ago when, for reasons unknown, they reverted to producing much simpler ones.
We tend to think of technological evolution as an exponential curve that starts out more-or-less flat in the early Stone Age and accelerates towards the present. But the idea that we are becoming ever more inventive may be an illusion. Looked at under the magnifying glass, the apparently smooth curve breaks up into a frenetic series of advances, retreats and new advances - what Peter Richerson, who studies cultural evolution at the University of California, Davis, describes as evolution "noodling about". In fact, over the whole of human history, we have probably lost more innovations than we now possess, says anthropologist Luke Premo at Washington State University in Pullman.
It is a sobering thought. Just when we were pinning our hopes on producing hi-tech fixes for today's problems - climate change, overpopulation, emerging infectious diseases and so on - comes the news that we are not advancing inexorably towards technological Nirvana after all. Nevertheless, a better understanding of how technologies evolve could hold some valuable lessons for the future. In building a more fine-grained picture of human technological history, we may identify clues as to what will work and what will not.
One of the long-standing mysteries of human technological evolution is why our Stone Age ancestors apparently showed so little inventiveness in their toolmaking. The oldest tools discovered to date are 2.6 million-year-old stone flakes in what is now Ethiopia. They mark the beginning of a refining process that didn't culminate in really effective stone hand axes until about 2 million years later. This slow progress, the flat part of the technological evolution curve, has been put down to the limited cognitive abilities of early hominins. Unable to learn from previous generations, each one had to start again from scratch, which explains why they lacked so-called cumulative culture.
Generally considered to be what separates humans from other primates, cumulative culture rests on two key skills: social learning, which is the transmission of knowledge to new members of a group, and over-imitation - the high-fidelity copying of a behaviour, including irrelevant or incidental elements, which allows the behaviour and its context to be passed along together. Some researchers have argued that cumulative culture only made its appearance around 100,000 years ago, with Homo sapiens (New Scientist, 24 March, p 34). But anthropologist and stone-tool expert Dietrich Stout of Emory University in Atlanta, Georgia, has challenged that view.
Innovation tends to happen by the introduction, deliberate or otherwise, of copying errors - the equivalent of genetic mutations in biological evolution - with those that provide an adaptive advantage being more likely to be passed on. Early humans might well have had what it takes cognitively to learn from their forebears, Stout argues, it is just that with the simple tools at their disposal, there wasn't much room for copying error (Philosophical Transactions of the Royal Society B, vol 366, p 1050). Put simply, he says, "you can't change much about a hand axe if you still want it to perform all the functions that a hand axe performs". However, as tool complexity increased, the potential for innovation grew.
Premo suggests another reason why Stone Agers' creativity may have been underestimated. Throughout those apparently uneventful 2 million years, they were hunter-gatherers who lived in extended, itinerant family groups of between 20 and 40 adults, plus children. "These small groups could have been exposed to fairly high chances of the whole group going extinct," he says, whether because their best hunter was incapacitated due to illness or injury, or because environmental conditions changed rapidly. When a local population died out, all its innovations would have died with it, and sometimes that could have meant the loss of generations' worth of know-how.
In 2010, with anthropologist Steven Kuhn of the University of Arizona in Tucson, Premo developed a computer model that recapitulated the behaviour of innovative, tool-using hominin groups in a Stone Age landscape. It showed that a period of technological innovation followed by the wiping out of the innovators and their kin looks identical, on the broad scale, to a period of technological stasis (PLoS One, vol 5, p e15582).
But if Stone Age toolmakers were innovating, then where is the evidence? The archaeological record is notoriously patchy and the further back you go the sparser it becomes. Even so, the apparent lack of progress may be partly explained by the timeframe we choose to consider. In soon-to-be-published work, Charles Perreault at the Santa Fe Institute in New Mexico gathered information about 500 archaeological samples - tools, pots and other artefacts dating from the past 10,000 years and coming mainly from North America - and analysed how they changed over time. He found that the rate of change depended on the period over which he calculated it, appearing to be rapid over short time periods and slower over longer ones. A key reason for this is that there are many advances and retreats over the shorter term that tend to cancel each other out in the longer timeframe.
There is an intriguing parallel here with biological evolution. Back in 1983, University of Michigan palaeontologist Philip Gingerich studied how shape and structure changed over millions of years in a wide range of animals. He, too, found an inverse relationship between rate of change and period of measurement and, like Perreault, concluded that this is simply an illusion of perspective (Science, vol 222, p 159). The main difference between the two studies is that, by Perreault's calculations, technological change happens approximately 50 times faster than morphological change.
As well as challenging preconceptions about the inventiveness of our Stone Age ancestors, these findings have also fuelled a growing realisation that technological innovations are highly prone to extinction. Premo and Kuhn's model hinted that there are many reasons why even seemingly clever inventions don't catch on, or die out. In the real world, a classic example can be found on the island of Tasmania. About 12,000 years ago, as temperatures and sea levels rose at the end of the last ice age, Tasmania was cut off from the Australian mainland and its inhabitants marooned. Archaeological evidence shows that until the land bridge was severed, Tasmanians possessed a range of complex technologies, including cold-weather clothing, fishing nets, spears and boomerangs. When Europeans arrived 10 millennia later, almost nothing remained. They found people whose technology was the simplest of any known contemporary human group.
Low population density and fragile networks for knowledge transfer were the main reasons for this loss, according to Stephen Shennan, director of University College London's Institute of Archaeology. He notes, though, that in other places and eras different influences have been at play. For example, market forces and political or social factors can dictate rates of innovation. A wealthy elite may be essential to sustain a community of craftspeople who need a long training period to learn to make the artefacts the elite desires. Patents, in the modern sense of the word, were invented in the 15th century, before which craftspeople found other ways to profit from their knowledge for as long as possible - ways that influenced the development of the technologies in question. Guilds emerged to protect skilled knowledge, for example, keeping the price high but the pool of knowledge transmitters small, and hence vulnerable to extinction if conditions changed.
Factors intrinsic to a technology may also determine its evolution. An example of this is found in Japanese katana or samurai swords, which remained unchanged for centuries because errors in forging the blades became too costly, discouraging experimentation. "We tend not to consider cost-benefit ratios," says Shennan, but they can be crucial. "Something that seems like a thoroughly useful innovation may actually disappear because of the costs associated with it." Conversely, a technology may spread at the expense of better alternatives because once established it is too expensive to change tack. An example is the QWERTY keyboard, which is slower to type on than other keyboard layouts, but continues to monopolise the keyboard market in English speaking parts of the world.
Rumour and gossip can shape the trajectory of a technology too. In the past, using a new tool or medicinal herb might have got you branded as a witch, encouraging people to hide or suppress discoveries. Religious institutions still have a special kind of power: by attaching moral or spiritual value to an innovation, they can usher it in, by denouncing it they can prevent its spread.
So, what of the future? Are things different now, enabling technological evolution to continue at an ever-faster pace? Because of the sheer numbers of us on the planet, sparse populations and fragile transmission networks no longer pose a serious threat to innovation. Besides, since the invention of writing, we have been able to store knowledge outside people's heads and disseminate it widely. But we may have unwittingly introduced other brakes on progress.
According to Alex Mesoudi, an evolutionary anthropologist at the University of Durham, UK, technological progress - as measured by indicators such as the rate of scientific publication and patents filed - has indeed been accelerating exponentially over the past few centuries, but is now showing signs of slowing. The trouble, he says, is that we have accumulated so much knowledge, that young people now spend proportionately more time learning from previous generations and less time innovating. Schoolchildren and students tend to learn a subject in the order that it developed historically. For example, physics undergraduates are tested on their grasp of pre-1900 discoveries. "Only at master's level do they start learning 20th-century stuff," he says. And that lag is having an impact.
In a paper published last year, Mesoudi pointed out that the mean age at which Nobel prizewinners made their prizewinning discovery, or inventors came up with inventions that were considered worthy of entry in prominent technological almanacs, increased from 32 in 1900 to 38 a century later. It is in this period that he found a decrease in overall rates of innovation (PLoS One, vol 6, p e18239). "There is some evidence that fields are slowing down," he says.
Something else is happening too. As technologies become more complex, the associated contextual or causal knowledge is being lost. People who build cars today do not necessarily understand how a car works, for example, since they may just assemble one part or operate a robot that does it for them. In Fiji, where houses have to withstand hurricanes, anthropologist Robert Boyd of Arizona State University in Tempe has found that locals have a pretty good grasp of why certain materials are better at withstanding hurricanes, but not why certain structural designs work and others do not. "Causal understanding is a very powerful and beneficial thing," he says. "If you are put in a different situation, due to environmental change, say, you can adapt much more quickly if you understand how a technology works than if you have to adapt as a population by trial and error."
It is not yet clear how much of a problem this is, since the information tends to be recorded and the body of people who do understand it, while relatively small, is probably still large enough to ensure preservation. In publishing his findings, Mesoudi intended to be provocative rather than pessimistic. He wanted to make people think about how ever-adaptable humans are adapting to the new problems that technological prowess presents. He suggests, for example, that one way we are overcoming the problem of that long learning period is through the collectivisation of science. What used to be a predominantly individual activity is now increasingly the occupation of groups who pool their knowledge. And there are potential benefits if it allows us to harness the power of the hive mind.
Mesoudi hopes that by building such adaptations into models of technological evolution, researchers will be able to make more accurate predictions and identify the factors that predispose an innovation to success or failure. Not all those factors will be under the control of innovators. Nevertheless, with better insight, they may at least be able to minimise the likelihood of repeating the experience of poor Zheng He, who lost the greatest fleet the world had ever seen.
Laura Spinney is a writer based in Lausanne, Switzerland