OLPC History: Senegalese Failure in Implementation

   
   
   
   
   

What does a Senegalese technology implementation project from 1982 have to do with One Laptop Per Child? Well, you might be surprised. At the same time that the French government was launching their successful (but quickly overshadowed by the Internet) Minitel project, they were also supporting a constructivist-based computer-learning project, using Apple II computers with the LOGO programming language/learning tool.

apple ii
Old skoll Apple II

Le Centre Mondial pour l'Informatique et Ressource Humaine, provided Apple II computers and the LOGO programming language to schools near Dakar, Senegal under the direction of Seymour Papert and Nicholas Negroponte. In the MIT Technology Review magazine, No. 13, May/June 1983, Dray and Memosky's "Computers and a New World Order" article reviewed the project:

The Center intended to use microcomputers to take computing to the people through educational workshops in both the developed and the developing world. Field projects were set up in France and Senegal... It was to be an international research center independent of all commercial, political, and national interests.
One Laptop Per Child trumpets this experience in its Progress page of its website.
In a French government-sponsored pilot project, Papert and Negroponte distribute Apple II microcomputers to school children in a suburb of Dakar, Senegal. The experience confirms one of Papert's central assumptions: children in remote, rural, and poor regions of the world take to computers as easily and naturally as children anywhere. These results will be validated in subsequent deployments in several countries, including Pakistan, Thailand, and Colombia.

S. Papert & N Negroponte
Now while the children took to the computers, the lab itself didn't fare well. The Technology Review conclusion after just a year from product launch?:
Naturally, it failed. Nothing is that independent, especially an organization backed by a socialist government and staffed by highly individualistic industry visionaries from around the world. Besides, altruism has a credibility problem in an industry that thrives on intense commercial competition.

By the end of the Center's first year, Papert had quit, so had American experts Nicholas Negroponte and Bob Lawler. It had become a battlefield, scarred by clashes of management style, personality, and political conviction. It never really recovered. The new French government has done the Center a favor in closing it down.

Is this One Laptop Per Child project fundamentally different than the first, a whole new program destined for success? Or is it simply an attempt to try the same thing again, this time with upgraded hardware (the custom-designed OLPC instead of Apple II computers) and software (using the programming languages of Scratch, a successor to LOGO, and Python, instead of LOGO) on a global scale?

Related Entries

28 Comments

I wanna know when angolan people will use the amazing technology of olpc?
We are waiting for you.

Important decisions in education in the past and present time have almost exclusively been based on value systems, ideologies and religion not on scientific knowledge. And lets admit that selecting what your kids should learn will always be linked with your own value system. Educating kids means shaping the future. And everybody wants a hand in that.

The challenge in OLPC's mission is not so much technology but to separate the learning goals which are entirely the countries' business from the learning method. Even keeping the learning method out of political hickhack is difficult because already creating higher numbers of independent thinkers is uncomfortable for many authorities be it of administrative, political or religious character. Yes, they should learn to use computers and to solve complex problems but they should not necessarily learn to ask difficult questions to their authorities let alone finding their own solutions to those questions.

OLPC's challenge is becoming even more difficult by the fact that constructionism to be effective needs adjustments in the learning goals. This will draw the attention and involvement of all interested parties and this is nearly everybody.

I guess that's why OLPC so far understandably refused to discuss curricula. However, not discussing the necessary adjustments of curricula is harmful to the effectiveness of constructionism. It is a dilemma. OLPC's approach probably has been their Trojan horse tactics trying to introduce laptops in large numbers disguised as e-book-readers ostensibly needing no adjustments of the education systems. After the laptops would be installed without opposition the growing debate about constructionism would create a large supporting bottom-up movement of kids and their parents opposing top-down forces from the mentioned suppressing authorities.

However, this tactics has probably already failed since the large scale orders have not been placed most likely because of the involved financial and educational risks. The necessary pilot tests will bring about the political discussion that OLPC probably wanted to prevent.

Now the only solution is to help keep the inevitable debate in the target countries as factual and informed as possible. Projects without any supporting facts will be beaten by those with questionable supporting facts which again will be beaten by those with solid supporting facts. This will be much more decisive than hardware features of the laptops.

"Important decisions in education in the past and present time have almost exclusively been based on value systems, ideologies and religion not on scientific knowledge."

You forgot one other important reason: leap of faith. Something like: "you have to take my word, this product will make wonders in your country/school/community".

Does it?

I have to say that after reading the MIT Media Lab India and Ireland backstories, I did not come away impressed with Negroponte's abilities in implementation: http://www.olpcnews.com/sales_talk/countries/india_rejection_back.html

Now that Jon dug up this review of his actual work with kids/computers in the developing world, I am even more suspect when I hear his loud "trust us" on OLPC implementation.

Oh wait, its wasn't a "trust us" it was "magic & miracles"! http://www.olpcnews.com/implementation/plan/implementation_miracle.html

Wayan,

Given the less than stellar implementation plans of previous projects, this is why testing is terribly needed. The countries involved should demand this. Someone will tell me again that testing is neffective/expensive, and the OLPC XO is so good that it can't fail.

In other words, a leap of faith, which is OK to me (it's not my money ;-). But if the program fails, who's to be held accountable?

Wayan,

"I am even more suspect when I hear his loud "trust us" on OLPC implementation.
Oh wait, its wasn't a "trust us" it was "magic & miracles"!"

Which brings us back to the decisions authorities in developing countries will have to make and your views on Intel vs. OLPC implementations - so let me ask again:

I think all here will agree with you when you say *"If you just look at the technology, the OLPC is far superior". Classmate PC, at $425 (less if you want to buy 1 mln units according to Luis Ramirez's post above) it's also much more expensive. So as far as hardware price and the usebility it gives a student, Classmate PC seems indeed a very poor value for money...

However, after reading so many posts of yours focusing on alleged or true lack of OLPC implementation methodology you stated *"Intel's World Ahead implementation methodology is just that - a world ahead of the One Laptop Per Child's ".

I read Intel's website but, apart some general statments about how great Classmmate hardware is and how Intel will bring education to the 3rd world, it doesn't give any detail at all.

In your "world ahead" comment you don't give any details either. Could we please know what the details of the World Ahead implementation methodology/plan is and what makes you think it is (or will be) successful ? A detailed critical review would be very much appreciated - thanks.

*Implementation Plan Challenge: OLPC XO vs. Classmate PC
( http://www.olpcnews.com/countries/brazil/olpc_xo_vs_classmate_pc.html )

This post rises a lot of interesting issues, most of them regarding continuity and resilience of Mr. Negroponte's approaches to development. Up to some point, it confirms some of my concerns about the "prophetic" nature of his, and makes the "Presidential strategy" more sensible, since Presidents and sometimes ministers tend to see projects from a wider perspective (or a less focused one, if you like). What may look brilliant from a President's office turns into an implementation fog at the minister's level, even more so at lower levels of officialdom.

Recently, most of the posts here have had a sort of "deathmatch" quality, with head-to-head comparisons intent on making crystal-clear that OLPC has the better computer. While this may be more or less evident, the issues about exactly what's the actual usefulness of the computer at primary education level is still to be properly addressed. I do believe it is a matter for each country; but at the same time, it is possible that there's a fundamental disconnection between our discussion and the realities on the ground:

What if the actual countries making the investment arrive to the conclusion that they don't really want to spend on computers for primary schoolchildren?

Thus far, it does look like that. To conclude: it is not a matter of which is the better machine, but what exactly any machine is good for at the educational level it is intending to serve. It may come to be that there is no good set of reasons (please, note the word "set") justifying the investment.

"To conclude: it is not a matter of which is the better machine, but what exactly any machine is good for at the educational level it is intending to serve."

There's no abstract concept of "better". The question is "better" for what?

On the very basic educational level, even if the laptops were initially to be used primarily for accessing content, whether eBooks (hope everyone agrees books ARE good for children's education...) and the Internet (live or cached like included Wikipedia) a choice of a tool (laptop) does make a difference. The choice of tools does determine not only how we can achieve certain activity but also whether the activity is possible at all.

If the tool can be easily damaged (dropped, water logged, worn out (fan), infested by insects...) or provide limited usability (poor readability, poor battery life) than, clearly, it's not suitable.

To conclude: given different offerings and limited funds, it's shortsighted and irresponsible to purchase a tool which not only will not be durable but also provide limited usage. As Wayan said elsewhere, "If you just look at the technology, the OLPC is far superior" and the "far superior" technology OLPC XO offers not only applies when used to access content - it's only, of course, the beginning of what is possible...

delphi,
in case the target countries don't want to buy laptops for primary schoolers then the XO, designed exclusively for that purpose, can be as good at that as you want and it will still not sell. If that is so then OLPC has missed the market or failed to create their own market.

The Classmate is less focused and could therefore be used for other target groups than primary schoolers. That might give it a decisive advantage although it is technically inferior.

Roland,

"XO, designed exclusively for that purpose, can be as good at that as you want and it will still not sell"

OLPC program does target primary school students. However, there's clearly nothing in XO hardware/software that makes it "designed exclusively" for that market. If the client countries decide to use it for secondary schools instead, its "fair superior" design will apply in that case as well - it's the conditions in which developing countries students are which makes XO a better choice and not the level the students are at. In fact I can easily envisage even many university students finding it more useful than anything else on the market...

delphi,
I agree with most you say. Peripheral changes would be sufficient to make the XO also a superior educational laptop for higher students. E.G. the Sugar GUI should be replaced by a windows based GUI e.g. Gnome. The keyboard should be standard size and the toy-like design is not necessary anymore. Additional storage could be added as memory sticks or memory cards.

If OLPC would modify their XO accordingly also primary schoolers could still use it by just loading Sugar GUI back on again. And what's more the XO could also be used for adults. Such a modified XO would technically dominate the whole field of low cost/low power computing. However, I don't know whether OLPC's sponsors would appreciate a low cost computer for adults creeping into the US market. AMD depends as much on a growing high-end CPU-market as Intel does.

"E.G. the Sugar GUI should be replaced by a windows based GUI e.g. Gnome. "

Sugar is windows based, just it doesn't use a desktop metaphore.

Please, can anyone explain to me why the Desktop GUI is superior to Sugar's setup?

For all practical purposes, I find it horrible.

It is heavily used in MS Windows, but that is because Windows' design is so bad.

Macs and Linux users generally use desktops only sparsely. The difference is in the much better design of the tool-bars and menues. As the XO uses a journal as a data store, even the hiearchical files systems are not needed.

Normally I use Gnome, but that is only to get at the tool bars.

Really, for normal (and collaborative) work, Sugar seems to be much better.

And rarely do I see a child working on a desk-top.

Winter

Rob Winter,
since a choice of GUI's can be made available without great technical difficulties there is no point in forcing anybody to use just one. The choice should remain with the user and nobody else. That's one of the ideals behind open source, isn't it.

"since a choice of GUI's can be made available without great technical difficulties there is no point in forcing anybody to use just one."

Sorry, I interpretted the use of "replace" and other remarks about Sugar as indicating that many think Sugar is a toy that should not enter serious schools (I remembered this thinking in the 1980s with respect to the Mac GUI). It certainly is not.

I seriously doubt whether the XO has enough memory to store two GUIs though.

Winter

I thought a lean version of Gnome would already be on board the XO. Maybe I am mistaken. Does any of our peer bloggers know more about this?

In case storage space would be the restriction for more than one GUI in parallel then there should at least be the possibililty to install only one from a choice of several. For adults I am anyway of the opinion that storage extensions like memory sticks or cards are necessary. Then also more GUIs could be offered.
Is the current network hardware/software of the XO also capable to connect to standard WLAN like 802.11/b,g ?

Winter - this week's OLPC community listserve has this interesting tidbit:

"Our Xorg 1.3 porting effort is progressing; it is semi-usable now on Bernardo Innocenti 's desktop. Input rotation has also been seen to work, but only for a brief lucky moment. We still have bugs to fix, but Adam Jackson is already starting to package things for us so that we'll be able to move to 1.3 consistently with the F7 upgrade.
"

So I'm not totally sure what that means; I thought Sugar ran on top of X already, but this could enable at least some of the lighter desktops (XFCE, for example) to run pretty easily. I ran a stripped-down KDE on a laptop with worse specs (excepting HD space) than the OLPC for almost a year before someone pointed me at XFCE; it worked, but was a bit klunky.

I mistrust Nick's belief in 'testing' almost exactly as much as I do the notion of magic and miracles.

It's had to overstate how much nonsense has ben foisted on the world by Rigorous Scientific Testing --- a.k.a "some selected data we have collected for you, see Tables 1 to 43, which prove conclusively whatever case we wanted to make, no matter whether that case is relevant or not to what we'd really like to know . . ."

I write ironically and even cynically, but my intention here is deadly serious. In the case of "Are computers useful in educating children?", I ask you to consider how we might begin to set about answering such a question, even given the best and most unbiased of motives?

Cheers,

Martin (Killer of Ninety-Nine Per Cent Of All Household Germs, Did You Know There Are More Germs On Your Kitchen Chopping Board Than There Are On Your Lavatory Seat?)

I take from your comment that you - being better informed than most of us - cut your steaks on your lavatory seat :-) What cooking do you do in the bowl? :-)

Martin,

About your eReader from your previous posts - although I like the idea very much, given the cost of the XO screen itself I thought your estimated figure was too optimistic and since the eReader purpose it to display static text and images only, the XO's screen would also be an overkill. Here is a much more practical, I believe , approach:

THE $10 E-READER PER CHILD PROJECT
( http://wikibox.googlepages.com/home )

Wikipedia - The ten dollar e-reader
( http://en.wikipedia.org/wiki/10er )

BTW. Even with my (computer) science background, I share your concerns about the validity of a 'scientific' tests in such a complex environment as education and it's especially so given the particular political and socioeconomic issues that it involves...

Delphi:
"given different offerings and limited funds, it's shortsighted and irresponsible to purchase a tool which not only will not be durable but also provide limited usage."

Again: this is predicated on government's deciding that their educational systems require a computer.

What if they don't? To elaborate: what if they decide that given the investment required, the gains do not justify to provide a computer to every kid?

Roland:
"Peripheral changes would be sufficient to make the XO also a superior educational laptop for higher students."

I've no idea how much will it cost to change the XO into an "older" computer, but it may not be trivial or peripheral. Also: there may be a lot of good reasons for schools to provide a computer that is similar to the ones kids are going to find in their potential workplaces; or at least, they might find it more reasonably to do so.

Eduardo,

"Again: this is predicated on government's deciding that their educational systems require a computer."

Correct. In which case the fact that XO is "far superior" than Classmate PC or other offerings is irrelevant... as is most of the discussions here on OLPC News as far those decision makers are concerned.

"I've no idea how much will it cost to change the XO into an "older" computer, but it may not be trivial or peripheral...there may be a lot of good reasons for schools to provide a computer that is similar to the ones kids are going to find in their potential workplaces"

Why would you want do that? Again think about the laptop as an educational 'tool' - do books these kids read now look and have the same content as the books/documents they are going to use in their future workplace...?

"Also: there may be a lot of good reasons for schools to provide a computer that is similar to the ones kids are going to find in their potential workplaces; or at least, they might find it more reasonably to do so."

This is an extremely dangerous route in education.

In the 1980s this was used in my country as an argument against using Macs in school. "Parent organizations" claimed their children should learn to use computer systems that prepared them for future work. And everybody knew that the future was not Apple Macintosh, but MS Dos.

Micrsoft software is infamous for the short shelf live of their knowledge. Whatever you know about the current incarnation of Windows, will be made useless by the next upgrade. Compare Windows 95/98/ME with NT/2000/XP/Vista.

On the other hand, a lot of what you might have known about Unix in the 1980s is still useful in Linux 2.6

Winter

hello delphi

What really grabbed me about Mary Lou's screen was that it looked like a brand-new technology with -- in particular -- ultra-low power consumption. I've looked around for other low-wattage screens but found none so far in the same general class.

Once we've dumped the keyboard-and-mouse -- we can still ask yes/no questions, since we have to have four key-switches anyway and can jump to different 'next page' according to the choice made -- I do truly believe that a bit more R&D aimed solely at lowering mean wattage while throwing away three quarters of the program and system overhead 'cause all we want is to bang pixels up on a screen, we might be heading for a $10 machine or close to it, in billion-unit quantities? Dunno, maybe I'm being a bit too optimistic (it's been known).

Pretty indestructible, though, virtually throwaway too, and very little temptation to sell it or swap it . . .

Cheers, Martin

Rob and Delphi: I agree with you that having a "business" computer standard when judging whether to have a computer in the classroom is silly / dangerous / preposterous. I'm just expressing a common view I've seen in schools, even at the primary level, in many countries. This particular point of view may reappear when decision time arrives. I oppose it and have tried to say so whenever asked about it.

But, there are too many still believing it.

Jon,
Thanks for digging up this article on the Senegal experiment.
Is this article available in digital format anywhere?
-M

The online pickings are slim. You can see part of the full quote on Google Scholar (search for Dray and senegal, the result is in page 7 of Artificial Intelligence and Education By Robert Walter Lawler, Masoud Yazdani: http://books.google.com/books?hl=en&lr=&id=iCbbXxnpM0UC&oi=fnd&pg=PP11&dq=senegal+negroponte+Dray&ots=PIqpKULR0l&sig=eWdVFK4CY5U1W6ysTlFeW5lJrkY

More on the 'Centre Mondial Informatique et Ressources Humaines':

http://www.euchner.us/papert.htm

The parallels to today are fascinating!

I note that Raj Reddy was also associated with this centre; John Markoff in the NY Times (16.8.04) suggested that his time there inspired the PCtvt. More info:

http://www.financialexpress.com/fe_full_story.php?content_id=78816
http://www.naute.com/heart/pctvt.php

Close