RAPUNSEL - learning programming as a first language

I'm at ISEA in Helsinki, where Mary Flanagan and Ken Perlin are presenting RAPUNSEL, a project where they're getting 11-13 year old girls keen on computers by - in collaboration with girls in the age group - designing a system where the kids program animated characters to choreograph a dance. It looks like a great project, and the manifesto has some excellent points:

If you're a kid, showing and telling things to the smart pets that you share with your friends is very different from writing Java, Python or Logo. It is much more powerful, because it builds on innate social and perceptual skills.

It is programming as a first language.

During the talk, Mary Flanagan pointed out that while middle school girls are as interested in and as good at technology as their male peers, only 7% of programmers in the US are women. That's why this project is targeting girls at this age. She's optimistic that games are a way to engage girls in programming, and she quoted statistics that apparently show that women account for 43% of all computer gamers in 2004 (39% in 2003), that half of all game purchases are by women and that 1/3 of Everquest players are women.

So far, they've been developing believable, animated characters that the girls actually like (the girls they're working with have very clear ideas of what they like and don't like) and they're setting up a system where kids actually program the behaviours of the characters. It looks a little like the Sims, but you can control the characters. Sounds like a great idea, and something that many kids would find interesting and empowering.

They've found a vast variety in taste among the girls, but generally, they want to fight. All of the kids in this study say Grandtheftauto is their favourite game. They don't always do what you'd expect them to do with the game - one girl just likes driving in Grandtheftauto, for instance. But still, girls liking violent games and killer robots (these girls don't like cute robots) is not what Brenda Laurel found in her work with girls and games. It changed a lot, though, from time to time, and also the girls' answers depend on whether they're alone, whether there are boys present, who's interviewing them... There's also a difference between the two groups of girls - I don't know areas of Manhattan so can't remember where, but in one group half the girls are living in shelters, in the other they're better off.

Mary also said that they're working to encourage social program. She said many studies have shown that peer programming is popular with girls but this is not being used in instruction - schools privilege individual work. I wasn't aware of that, and I'd love to know whether readers here have that impression too - do women prefer peer programming? Are you aware of work on this?