TOOLS MAKESTH THE MONKEY

原文链接：http://www.newscientist.com/channel/life/mg20026772.100-tools-maketh-the-monkey.html

Teach them to use tools and monkeys start behaving in oddly human ways. It^’s like they^’re following in our evolutionary footsteps, says Laura Spinney

Pin is no ordinary monkey. If you stick out your tongue at her, she^’ll return the compliment. If you show her how to open a latched box, she^’ll rotate the box until she finds the latch, then open it herself. Place a piece of fruit just out of reach and she^’ll pick up a rake and drag the fruit towards her. All of which is remarkable, because monkeys are not natural imitators and nor, as a rule, do they use tools.

Pin is a Japanese macaque trained by Atsushi Iriki, head of the Laboratory for Symbolic Cognitive Development at the RIKEN Brain Science Institute in Wako, Japan. According to Iriki, Pin^’s remarkable abilities offer an unprecedented window on one of the most important events in evolution: the emergence of human intelligence.

It^’s well known that non-human primates can be taught to do very human-like things: chimps and orang-utans have wowed the world with their ability to learn sign language, for example. But the general assumption is that there is an immutable upper limit to their abilities. Iriki thinks otherwise. ^“It is often said that a monkey has the ability of a 2-year-old kid, and an ape that of a 7-year-old, but it^’s not like that,^” he says. ^“Through training, you can extend the spectrum of the monkey^’s ability closer to a human-like intelligence.^”

Iriki is not attempting to raise a race of supersmart monkeys. He is much more interested in finding out what macaques can tell us about ourselves.

His controversial hypothesis is that the macaque brain contains all the components that gave rise to human intelligence – it^’s just that they have never been assembled. In the macaque^’s natural environment, there was no demand for it. But put the macaques in a new environment, place new demands on them, and you can unlock that potential. This, Iriki says, is the nearest thing we have to experimentally rerunning the tape of human evolution to reveal how our primate ancestors went from leaping about in trees to writing symphonies and sending spacecraft to the moon (Philosophical Transactions of the Royal Society B, vol 363, p 2229).

Iriki^’s starting point, like that of many others before him, is that the explosion of human intellectual abilities was triggered by tool use. When primates stood up on their hind legs, they freed up their hands to use tools. This somehow unleashed a cascade of brain changes that led, slowly but inexorably, to the evolution of higher cognitive powers such as self-awareness, language and intelligence.

The tool hypothesis has been around for many years, but exactly how tools kick-started a revolution remains unclear. It is also at odds with the fossil record. The earliest known stone tools are about 2 million years old, yet nothing resembling fully intelligent behaviour emerged until 200,000 years ago. ^“The idea that you start using tools and, bang, all the rest follows – that^’s not what the fossil record seems to suggest,^” says social neuroscientist Christian Keysers at the University Medical Center Groningen in the Netherlands.

Iriki^’s unique perspective on the problem is that tool use was the catalyst for a much more important mental breakthrough, albeit one that took 1.8 million years to unfold: the emergence of a sense of self. By this he means the ability to conceptualise one^’s own existence in time, plan for the future and understand ^“intentionality^” – your capacity to change your environment.

So how did tool use give rise to a sense of self? Iriki believes the starting point is the way tools induce a modification of body image – the basic mental representation of ^“self” that consists of knowing where the physical body ends and the environment begins. When we use tools such as hammers or tennis rackets, we integrate them into our body image;

“The changes in the monkey breains show provocative similarities with humans”

Our brains treat them as a temporary extension of our hand orarm.To turn a stone or a stick into a tool,our ancestors would have to have done the same.This ,Iriki argues,led to the gradual dawning of a sense of self more so phisticated than the basic body image,creating a new evolutionary force that rapidly ratcheted up intelligence.“Once you have a sense of self,you can intentionally control the environment,and that modified environment in turn puts selection pressure on your brain,”Iriki says.He has dubbed this dynamic,two-way interaction between brain and environment”intentional niche construction”,and argues that it is the missing link in the story of human evolution(see “Home-made evolution”,below)

Sense of self was crucial for another reason:it allowed our ancestors to conceive of the existence of other selves,each with their own intentions.This is the essence of “theory understading and hence communication,language,society and culture.

So far so philosophical,but Iriki’s ideas are more than just speculation.They are the culmination of years of experiments designed to see how far hecan “evolve”macaques towards human-like behaviour by training them to use tools.

In one series of early experiment,published in 2000,Iriki and colleagues trained Japanese macaques to use a rake to retrieve fruit placed out of their reach (Canadian journal of Physiology and Pharamacology,vol78,p958). It took the macaques about two weeks to master the technique, but once they had, Iriki found their new skill had left striking traces in the brain.

During training he and his team recorded the electrical activity of neurons in the parietal cortex, the part of the brain thought to encode body image. These neurons respond both to touch and to visual images of the monkey^’s own body, and are thought to meld the two into the body image.

At first, the rake did not trigger activity in the parietal cortex, even when the macaques were holding it. Over the training period, however, the neurons gradually started to respond to the rake, and by the end of the training were firing in response to it just as they would to an arm or a hand. In other words, the macaques had incorporated the tool into their body image.

In the next set of experiments, rake­ trained macaques were presented with a tougher task: to retrieve food from under a screen that obscured their hand, the rake and the fruit from view, guided only by video images on a screen. Before long they were raking up fruit as skilfully as before. Now, the parietal neurons were responding to the video image of the rake-wielding hand.

Home-made evolution

What this suggests is that, given the right training, the macaque brain is capable of modifying its body image to incorporate a tool and can mentally displace that image onto a screen and still recognise it as ^“self^”. Iriki claims this means his trained monkeys have a body image as sophisticated as that of a 9-year-old child – not quite a full-blown sense of self, but more sophisticated than your average monkey.

Training can also unlock what may be the rudiments of language. In another set of experiments, Iriki and colleagues noticed that some tool-trained macaques started making cooing noises when they could see food but the rake was out of reach. Such noises are known as ^“reference calls^” – voluntary vocalisations referring to specific objects. Wild macaques never make such noises; their vocalisations are limited to reflexive responses triggered by external events, such as the appearance of a predator.

Iriki and his colleagues went on to successfully train two macaques to request food or the rake using cooing noises. Though each macaque^’s calls were different, Iriki says that it should be possible to teach naive macaques to imitate the calls of others, learning what they mean in the process. That way, the calls would acquire communicative power (Neuroscience Research, vol 45, p 383).

Another distinctly human-like skill that can be brought out by tool training is imitation and joint attention, which are the twin pillars to communication and the propagation of culture. In the wild, macaques show very little interest in one another; infants have occasionally been observed to follow their mother’s gaze and imitate her facial expressions, but adults never do.

This is puzzling, because macaque brains clearly monitor the behaviour of others. Macaques were the animals in which mirror neurons were discovered in the early 1990s. These are nerve cells which fire not only when the animal performs an action but also when it sees the same action being performed by another.

Iriki reasoned that the missing element in wild macaques might be a sense of self, and the related ability to appreciate the existence of other selves. He predicted that tool-trained monkeys might start following others^’ gaze (the essence of joint attention) and copying their actions, just as human children do.

Iriki^’s monkeys are already raised like children, in an intense relationship with a human carer, on the rationale that this may help them learn joint attention and imitation. Sure enough, he found that three out of four monkeys raised this way, and then trained to use tools, ^“got^” joint attention. Having achieved that, two of them – Pin included – began to spontaneously imitate an experimenter^’s actions, such as sticking out their tongue (International Journal of Psychophysiology, vol 50, p 81).

Iriki has also been tracking how his monkeys^’ brains change as a result of their training, and has found yet more provocative similarities with humans. In an as-yet unpublished brain-scanning study carried out with a team at University College London, his group has shown that rake training causes areas of the prefrontal cortex and parietal cortex to expand. In humans, the prefrontal cortex is important for controlling behaviour in complex social situations, and is known to have expanded rapidly during our evolution.

Retracing the steps

Taken together, Iriki believes his findings strongly support the idea that training macaques to use tools recapitulates at least some of the evolutionary steps made by our ancestors. He does not claim that he can elevate macaques all the way to human intelligence, but he believes they are a useful tool for probing how the raw material of the primate brain evolved into the highly sophisticated thinking machine in our heads.

Many of Iriki^’s conclusions are based on just a handful of trained macaques, and so most researchers would like to see more work before allowing themselves to be convinced.

Nevertheless, Keysers says he finds Iriki^’s ideas original and intriguing. ^“One of the most exciting findings is just how much of the brain connectivity can really be changed by this training regime,^” he says.

Others are more sceptical. Daniel Povinelli, director of the Cognitive Evolution Group at the University of Louisiana, Lafayette, says it is not surprising that training induces modifications in the macaque brain, but questions whether these changes recapture events in human evolution. There^’s no reason to suppose that the seeds of human-like intelligence are lurking inside the macaque brain just waiting to be unlocked, he says. ^“Iriki has this very progressive view of evolution – that somehow monkeys are incomplete human beings,^” says Povinelli. ^“That^’s just not the case.^” He points out that macaques have been evolving on their own trajectory since they split from the human lineage around 25 million years ago. They did not just freeze at a lower rung of the evolutionary ladder, he says.

^“That^’s not the point,^” Iriki counters. ^“What I am interested in is the common precursors, how one species could be induced to resemble another, and what kind of brain changes would accompany that.^”

Iriki^’s next move is to start working with marmosets, New World monkeys which can also be trained to use tools and which have the advantage of breeding more rapidly than macaques. They can also be genetically engineered, which raises the possibility of endowing them with traits that may be even more human-like. For now, though, Iriki is keeping the details of the work under wraps.

Aside from providing insights into our past, Iriki^’s work also gives tantalising glimpses of the future. One of the messages of his research is that the human brain is still a work in progress. Tool use may have started us on our journey, but there^’s no reason to suppose that we^’ve unlocked all the potential in our brains. Perhaps new ^“metaphysical^” tools such as computers and the internet are already taking us to the next level. ^“Mind once emerged in our brains,^” says Iriki. ^“What might emerge next?^”

Laura Spinney is a writer based in London and Paris

www.newscientist.com/channel/being-human/

本文地址（转载请注明出处）： 复制
收藏、分享这篇文章：        更多...