(This is a letter I sent to the editors of IT Week, a British publication, in response to the following commentary by Roger Howorth. I apologize to readers for continuing the OLPC thread when I said that I would be more constructive, but this offers a fairly clear opportunity to state my underlying objection to the project.)
Roger Howorth's enthusiasm for the OLPC (one Laptop Per Child) project is not supported by research, whether conducted by him or others. He speaks airly of "40 years of research by (project founder Nicholas) Negroponte and others", of a 1988 project in Costa Rica, and a "recent initiative with about 50 laptops in Cambodia".
I have been looking into the research underpinnings of OLPC and find them remarkably thin. Recently I asked a project member (Mihail Bleitsas, Chief Connectivity Officer) about the Cambodian example. He could not provide the name of the village (though he did send it to me by email later - it is Khum Reaksmey) and when pressed admitted that no research report of that experience exists. This is significant because Negroponte consistently refers to this example as seminal in the formation of the project, though he provides only the sketchiest anecdotes as justification.
In my discussion of the project on my blog <www.fonly.typepad.com/fonlyblog/2005/11/problems_with_t.html> I have received many responses to the effect that "I trust MIT to have thought this through". When talking with members of the project I am told "MIT has done research into education for 35 years, and that's enough - we're going to build the laptop". But nowhere can one find a compilation of the research in question or an argument as to how this particular solution will achieve its vaguely-stated goals of elimination of poverty through education - one is simply expected to accept on faith that this is the best and only solution to the problem of poverty.
We are thus confronted with a project whose basis is: "we will do this because we can and because we have the highest political and economic connections, and any questions about whether and how it will work will be dismissed." This project does not rest on a scientific foundation. If it did there would be research reports, discussion and argumentation on the topic in pursuit of the development of consensus.
I challenge anyone to find such a process of discussion and debate in the history of the OLPC project. For those who tell me how much they trust MIT, I respond that OLPC did not make the most elementary power calculations (which I enumerate in my blog) before announcing the signature hand-cranked power source for the laptop. A few days after I posted these calculations all references to power generation disappeared from the OLPC website. What is there to trust here? Mr. Howorth might look into this question rather than simply join in the cheerleading.