Tuesday, September 30, 2008

Robot abuse

Two BBC journalists have posted an intriguing video of an informal experiment carried out in what I guess is their journalistic HQ. The film shows them acting as mum and dad to a kind of robot dinosaur called a Pleo. I'm not sure what AI wizardry makes the Pleo's behaviour so lifelike, but it is certainly convincing. Of most interest to me, though, is the contrast the film draws between two parenting styles. The Pleo's manufacturers claim that the robot forms a personality on the basis of its early experiences, much as a human infant is thought to do. Under a loving, indulgent parental regime, the robot seems well-adjusted and contented. Under a harsh, abusive regime, it ends up lethargic and depressed. 

Plenty of interesting things emerge from this piece. Firstly, I'm struck by how good these journalists—we're not told whether they are parents themselves—are at playing the part of good parents. Instinctively, it seems, they do exactly what good parents everywhere do. We see evidence of motherese in their conversational styles: higher-pitched voices with exaggerated intonation. The Pleo gets gentle tactile stimulation and plenty of mind-minded linguistic input, concerned with the robot's emotional and cognitive states of mind. The 'parents' are keen to take the Pleo outside to expand its experiences. There is even a suggestion that these positive strokings give the Pleo the 'confidence' to go out and explore its environment: a hallmark of secure attachment behaviour. 

In the neglectful scenario, the parents play their roles in an equally, and scarily, convincing way. They try to pass the baby on to each other, making excuses for why they can't look after him. The poor robot baby is shunted around the office in the hope that someone will give him some time. 'He's a more reserved robot than in the parallel happy life,' the caption tells us. It might seem odd to attribute such human characteristics to something that we know to be a lump of electronic circuitry, but the emotional impression is powerful. 

As with so much of developmental psychology, the adults' actions are of as much interest as the baby's. The Pleo, of course, has been cleverly designed to press all the emotional buttons that a human baby does, with cute squeaks and ET-like eye-blinks. It reminded me of the wonderful Kismet robot created by MIT engineers:



The point about is Kismet is that it calls out certain reactions in us, just by behaving in a human-like way. One puzzle that developmental psychologists have to contend with is how parenting can possibly make any difference when the object of parental love—the newborn baby—seems to have so little to make a difference to. Here's a creature that—very broadly speaking—cannot see clearly, process information effectively, remember past experiences, have consciousness of its own self and so on. How can it possibly be shaped by such sensitive and subtle parental behaviours? The robot work gives us a clue how. A creature that starts off as a blob of unconscious circuitry might become conscious, just because we take it as being so. If that is so, then the Pleo's parents' efforts might not have been wasted after all. 

Friday, September 26, 2008

Doddy and Daddy

Angelique Richardson has also responded to Alison Gopnik's TLS review, mentioned in a previous post. She draws our attention to 'one fine child psychologist', Charles Darwin, quoting from his 1873 work The Expression of the Emotions in Man and Animals. Darwin was indeed a pioneer of infant observation, which was most memorably documented in his short article 'A biographical sketch of an infant', published in 1877. The subject of the article was Darwin's infant son Doddy (properly William), who was born in 1839. The piece contains many rich observations, but of most interest to me were Darwin's comments on Doddy's reaction to his reflection in a looking glass. Doddy's mirror reactions were first observed at four and a half months, at which stage he seemed to take his own reflection as the image of another being, a Doddy double. Later, he began to make the connection between the reflection and the person who was being mirrored. In The Baby in the Mirror, I describe how Darwin's observations inspired some of my own: 
Two months later, Doddy’s understanding of mirrors had taken a step forward. Facing the mirror in front of his father, he now seemed to realise that his father’s reflection was connected to the person standing behind him. When Darwin made a face in the mirror, Doddy turned to look at the man, not at his reflection. We saw Athena doing something similar at the same age. We would sit her on the bed and stand to one side, so that she could see us in the mirror but not herself. She would raise her arms and throw herself forward, reaching, as though in prayer. When we waved, she would give her characteristic double-handed wave in return, a cross between a communicative gesture and an exuberant attempt at taking flight, raising both hands high and then batting them down on the duvet on either side. Then she would turn to look at the figure reflected, confirming it against its image. She understood something about the mechanics of reflection, that what you see in the mirror is not just an extension of the world but a special version of it. She was far from having a full understanding of it, but the looking-glass world was becoming real.
You can read Angelique Richardson's letter here

Saturday, September 20, 2008

I'm just going to have to put you on hold

The arrival of hard-disc recording has transformed our television habits. No matter how enthralling the programme—try dragging Athena away from an episode of the Australian kids-soap Mortified, for example—a press of the button will freeze it in time and unfreeze it when it is safe for the action to resume. Usually the hiatus lasts only as long as it takes to communicate essential orders about tidying bedrooms or getting ready for bed. Crucially, the kids understand that their precious viewing can be manipulated in this way. They might not appreciate the technicalities of digital recording, but they understand that their glimpse of TV-land will be frozen there, incorruptibly, on the screen until such time as they can pick it up from where they left off. 

Isaac is on the phone to Granny. They are discussing some work he has done at school, and he wants to bring it to the phone. 'I'll just pause you for a second,' he says, putting the handset down and skipping off to retrieve his drawing. For a moment, Granny is just like any bit of digital reality that can be paused, or even rewound, for the user's convenience. Is he using the word metaphorically, in the sense of 'keeping someone hanging' or 'leaving someone to stew'? I'm not sure. I suspect there's a deeper confusion about the particular multimedia experience that's called interacting with a person. Or perhaps he thinks that pausing his phone conversation is about exerting control over the bit of technology he holds in his hands, rather than using his powers of persuasion on the flesh-and-blood person at the end of the line. 

All of which reminds me of a cartoon I was sent a while ago. A stout boy is standing on his back step pointing a remote control at his garden and clicking furiously. Out of view behind him, the voice of Mom is calling: 'It's the outdoors! The remote doesn't work on it!! Go and play...' I don't think we're meant to conclude that the little telly addict depicted here does not understand the outside world as a reality that exists independently of him. He knows there's a real world out there; he's just a bit confused, like Isaac is, about the mechanisms through which it can be controlled. 

Monday, September 15, 2008

Sticky mittens

Avid readers of the TLS will have noticed that it recently carried a review of The Baby in the Mirror by Alison Gopnik. You can read Gopnik's review here. I responded to the review and the letter was published this week. You can read my response here.

Friday, September 5, 2008

Packed off to the brain gym


'Is playing with my train set using my brain?' Isaac asks. 

I blame the video games for this current obsession. His activities now divide neatly into those that involve using your brain and those that don't. Colouring and looking at storybooks are good because they employ your grey matter; playing Mario and Sonic at the Olympic Games confers no such benefits. Ever since he discovered Athena's Nintendo DS (and refused to give it back), he has been a little too interested in things that beep, flash and emit electronic sighs of disappointment when you lose. 

I haven't yet quizzed him on what he understands about the body's most complex organ. A landmark study by Carl Johnson used an innovative method to get at young children's understanding of how our identity as people is underpinned by our biology. He presented children with hypothetical situations involving transplants of different parts of the body. For example, children were asked to imagine that their own brain were removed and inserted into the body of a friendly pig, or that there were an exchange of brain, mouth or heart between themselves and a baby. Would children understand that removing a person's brain and putting it into the head of another would affect the identity of that person, while an equivalent transplant of another organ would not affect the consciousness of either party? 

The results of this and other studies show that children's understanding of how mind is rooted in brain develops gradually over the early school years. Children of kindergarten age seem to understand that the brain is where thinking happens, and that without it consciousness would be impossible. It is a slightly different matter, and one that requires further developmental progress, to understand how the contents of our consciousness are 'contained' in the mental organ. Gail Gottfried and colleagues have argued that children's increasing acceptance of the 'brain-as-container' metaphor is at odds with their essentialism. This concept from philosophy, when applied to the development of children's knowledge, refers to the supposition that living creatures are what they are by virtue of some essential quality, such that a pig will remain fundamentally piggy no matter what modifications are made to its insides. Four-year-old Isaac might accept that his brain is what makes thinking possible, while insisting (because of his rampart essentialism) that he would stay the same person even if he were magically implanted with someone else's cortex. 

The need to maintain the processing efficiency of the porridge between your ears is becoming more obviously acknowledged in children's culture. I've caught Athena watching a programme called Brain-Jitsu on CBBC, which sets its child contestants brain-training challenges based on elementary neuroscience. Some of the functional neuroanatomy is a bit simplistic—there is too much of that left-brain/right-brain twaddle that was so popular in the 1960s—but it surely can't be a bad thing that kids' interest in the science is getting this potential boost. 

Anyway, they are now away at the brain gym to beat all others: school. Isaac was packed off to his first day at reception this week, with the promise that he would be doing lots of things that involved using his brain. Life at school will be like life at home when he's not using the Nintendo, only more so.  

The end of the school holidays, and the prospect of weeks of undistracted work time, coincides with the news of two polls that show worrying levels of stress-related drinking among parents of young children, especially at the end of the school holidays. Now there's a news story that rings true.