The Hole in the Wall

1969 My friend Gora takes me to the Odeon in Delhi to see Arthur Clarke and
Stanley Kubrick's "2001: A Space Odyssey". A black monolith is found
in Africa and another on the moon. The monoliths seem to do nothing at all. All
humanity is desperate to learn more about them. We love puzzles.

1987 I have a PC with a 10 megabyte drive. I spent Rs 50,000 to buy it,
nearly a year's salary. My son is six and I need to protect my PC from him. He
agrees not to touch it. But can he watch? Sure, I say. Let's see what sense he
makes of DOS commands.

A month later, I have lost a spreadsheet file. I search directory after
directory scrolling up and down the file lists. A small voice from the back
says, "If you do dir/w all the files will show side by side as well".
I gulp. "Really?" I say, "Well, you can try it yourself, just be
very careful."

Two months later he is playing Moonbugs, Simcity and Flight Simulator. Six
months later, he does things with DOS that I never thought were possible. He
types faster than I can.

1988 I write a paper for the annual conference of the All India Association
for Educational Technology. It says maybe we should just give PCs to small
children and leave them alone to figure it out by themselves. Surely, some smart
kids will learn to use it by themselves. Let the best get better, the rest can
follow later, says the last part of the paper. The delegates think it's a
fascist, preposterous idea.

1999, January 26
My friend Vivek has a computer sunk into a wall near our office in Kalkaji,
New Delhi so that its screen was visible from the other side of the wall. A
touchpad was built into the wall as well. Children came running out of the
nearest slum and stuck to the wall like glue. A few hours later, a visibly
surprised Vivek said the children were surfing.

A few days later, the media arrived and this little experiment got the entire
world's attention. I think every newspaper and every TV channel on earth
reported on what the media called "The Hole in the Wall". 

But where does this magical computer literacy happen? It took five years, a
lot of travel and a lot of money to find out. There were great surprises and
great disappointments on the way. But, in the end, Nature's lessons were
simple, direct, and, in retrospect, obvious.

(Left) Kalkaji, South Delhi, 1999: The original “Hole-in-the-Wall”; (Right) A “Hole-in-the-Wall” kiosk in Cambodia, a gift from the government of India. The Cambodian kiosks, the first outside India, are showing a computer-literacy impact similar to the ones in India

After the heady rush of the Kalkaji experiment, it was time for some
scientific introspection. Within six months, the children of the neighborhood
had learned all the mouse operations, could open and close programs, surf the
Internet and download games, music and video. When asked, they said they had
taught themselves. They were describing the computer in their own terms, often
coining words to describe what they saw on screen. The hourglass symbol was
"damru," the mouse cursor, "sui" or "teer".

We repeated the experiment twice, once in the city of Shivpuri in Madhya
Pradesh (Digvijay Singh was interested) and once more in a village called
Madantusi in Uttar Pradesh. The Madantusi experiment was the first time we got
external funding: Dr Urvashi Sahni, principal of a school in Lucknow and a keen
educationist, raised the money from NRI friends in the USA.

Both experiments showed the same results as in Kalkaji. The children seemed
to learn to use the computer without any assistance. Language did not seem to
matter. Neither did education. Had we stumbled on to something universal about
children and computers? I needed research funding, desperately.

The funding came in quite on its own! The ICICI Bank's Social Initiatives
department wanted to put ten computers in villages in the Sindhudurg area of
Maharashtra. A young banker called Bikram Duggal, to whom thousands of rural
children and I should be grateful, backed the idea.

Sheila Dikshit, the chief minister of Delhi, visited Kalkaji and the then
principal secretary, Regunathan, a long-time educationist, took what he called a
calculated risk. He funded 30 computers in the Ambedkar Nagar area of Delhi.
Today, more than six thousand children need to thank him for that decision.

And finally, James Wolfensohn, president of the World Bank, came to Kalkaji.
He had heard of the experiment from Peter Woicke, director of the International
Finance Corporation. Together, they decided to fund a three-year research
project to find out if what had happened at Kalkaji was a freak accident or a
universal learning mechanism. That decision would affect the futures of 40,000
children to begin with and, perhaps, that of all the children in the world one
day.

India is as good a laboratory as any for finding out if any child in the
world would respond in the same way to a "Hole-in-the-Wall" computer.
We have all the diversity in the world that can possibly be had between
children: economic, social, genetic, cultural, physical, you name it, we have
it.

India is also a good laboratory for finding out if computers can survive
outdoors, built into walls and buildings. We have all the different kinds of
weather conditions any other part of the world has.

The research design was simple. We would install computers in 22 locations
across India, chosen for their diversity of human and climatic conditions. We
would then choose 15 random children as a focus group and measure their progress
over nine months. We would then compare them with other children who did not
have exposure to computers and see if we could get to an explanation for the
"Hole in the Wall" effect.

But how does one measure computer literacy? We needed tools. We found tests
on the Internet that took hours to administer: too long for our purpose. The
funding clock was ticking and we had to find a solution fast.

Desperation can often work wonders in science. In this case, it produced an
idea.

Two scientists from our centre, Parimala and Ritu, proposed a test wherein
you are shown the 77 common icons in the Windows environment. You are then asked
to describe what they might be for. If your description is correct you get a
mark.

It seemed like a simple but stupid test. We all use computers quite well, but
do we know all the icons? Of course not: some of us don't use the icons at
all.

Pari and Ritu did not listen. They administered the test to all our office
people. The results were interesting: we got around 70% of the icons right,
irrespective of whether we used icons or not. Not only that, the people who did
a lot of spreadsheets got the Excel icons right, the ones who did a lot of word
processing got the Word icons right. It seemed as though familiarity with an
application gave you the ability to guess that application's icons correctly,
even if you had never used them.

Another colleague of mine, Monica, and I did a validation of Pari and Ritu's
Icon Association Inventory (the official name for the icon test). We gave two
sample groups the icon test, along with the much longer American test of
computing literacy. The results displayed an incredible 98% correlation. The
icon test measured in 20 minutes what the other test would in two hours. We had
our measuring instrument. Ritu added a battery of intelligence and creativity
tests to the icon test, our basic tool.

While we were sharpening our measuring instruments, Ravi Bisht, the creator
of India's first shopping catalogue application among many others, was
traveling from village to village all across the subcontinent. Wherever he found
an appropriate place and a friendly panchayat or school, he built a little
structure with three computers facing the road. His design is today the standard
design for the "Hole in the Wall".

Sanjay Gupta, head of our centre, had another set of problems to solve.
Touchpads would fail within weeks of installation outdoors. Key tops would
vanish. Air-conditioning would cost more than the computers. The dust would get
into everything.

In the years between 1999 and 2002, Sanjay and Ravi solved all these problems
through a series of inventions: new mice, keyboard covers, reversed exhaust fans
and thousand other vital little things. They made it possible for ordinary PCs
to work outdoors. Anywhere.

If a computer went into a hang in a remote village in Kanyakumari, would
someone have to fly from Delhi, just to press the reset button? We wrote
software that would enable us to "see" our computers from anywhere
through the Internet. We wrote software to prevent Windows from hanging. We
wrote software to protect our desktops from getting accidentally deleted.

And now, we were ready.

In 26 locations, with 100 computers standing in remote villages, our field
observers began testing. Focus groups were tested for nine months and the
results compared with control groups and other frequent users. An estimated
40,000 children used these computers. They have all made themselves computer
literate. The average icon test scores stand at 40% in nine months. We had our
proof for self-regulated learning. And this time we know that it will happen
anywhere in the world, to any child, in any climate.

I decided to call this method Minimally Invasive Education (MIE). The rest of
the world continues to call it the "Hole in the Wall". Amen.

In the meanwhile, the government of India gifted five kiosks to the
government of Cambodia. These were the first ones we would build outside the
country. The library of Alexandria in Egypt wanted to build 30 kiosks. We sent
them kits and detailed instructions. The CSIR of South Africa built two kiosks
in villages of South Africa. The results were identical to those in India.

The French embassy in Delhi sent Pascal Monteil, a well-known new media
artist from Paris, to three of our villages with playground computers. Parimala
went along with him to ensure that he did not "teach" anything. The
children observed and questioned Pascal intently. Then, they taught themselves
digital photography (using his camera) and Photoshop in French. Then they
started to draw. "This does not exist anywhere ... except in your mind and
the computer," said Moumita, an unkempt little girl from Bishnupur, 24
Parganas, West Bengal.

Pascal took back over a hundred digital creations. He said in a lecture that
his concept of how art should be taught has changed.

If the Hole in the Wall is changing the children's analytical skills,
English and certain value systems, then its consequences are far greater than I
could have imagined.

In the slums, local adults laugh. "If you take away their free time,
petty crime and other naughty stuff will not happen," they say. I am
amazed. Never thought of it that way.

There is a frequently voiced concern about access to pornographic material
through kiosks that are connected to the Internet. While this has happened once
in a while, it is quite rare. In the last five years we have had less than 1.5%
pornographic usage at any connected kiosk. This is because of several reasons.

First of all, our kiosks are meant for children below 15 years. That audience
has only some marginal curiosity about pornography, that too among the upper age
groups.

Second, our kiosks are in highly visible, public places. It is rather
difficult for children to browse pornographic content when they are in
heterogeneous groups and where the screen is visible to passing adults.

Third, our kiosks are monitored remotely and the screens are visible over the
Internet. The children know this, as every kiosk has a sign announcing this.

The few instances of pornographic access that the remote monitoring system
has recorded in the last four years have been due to access by adults. The
reason why such access by adults is not common is because of an interesting
design feature.

The kiosks are designed for use by children. The screens are placed lower
than the average adult's height so that you would need to stoop to see them. A
lid on top of the screen acts as a sunshade and that too tends make it difficult
for taller people to crawl under it. There is a cowl on top of the keyboard and
mouse that has a gap through which only small hands can go in. There is a
seating rod that is placed such that an adult seating on it would not have
enough legroom.

All this ensures that an adult needs to be in a rather peculiar position to
use the kiosk. It is interesting that some adults have, nevertheless, not been
thwarted in their attempts to misuse the kiosk. Fortunately, their numbers are
small.

A very large number of children seem to benefit from the kiosks. In
independent studies conducted at Madangir, New Delhi, three organizations
concluded that about 6,000 out of a total of 9,000 children in the area were now
computer literate. This was achieved over three years through 20 effective
computers in the Hole-in-the-Wall configuration. This seems to indicate that up
to 300 children can share one playground computer.

Based on the experience and data gathered over the last four years, we feel
that such "playground" access points should be a part of every primary
school. Where primary schools are not available, such facilities could provide
even more vital "emergency" educational inputs.

Minimally Invasive Education through public Internet kiosks for children
should form an integral part of primary education in the twenty-first century.
It can not only close the "digital divide" rapidly, but can also
unlock the creative potential for self-development inherent in children in ways
that educationists would not have deemed possible even a few years ago.

Dr Sugata Mitra is Chief
Scientist, Center for Research in Cognitive Systems, NIIT Ltd

What do they Learn? 

An estimated 300 children can learn to do most or all of the following tasks
in approximately three months, using the "Hole-in-the-Wall"
arrangement with a single PC:

• All
  Windows operational functions, such as click, drag, open, close, resize,
  minimize, menus, navigation etc
• Draw and paint pictures on the computer
• Load and save files
• Play games
• Run educational and other programs
• Play music and video, view photos and pictures
• Browse and surf the Internet, if a connection is available.
• Set up e-mail accounts
• Send and receive e-mail
• Chat on the Internet
• Do simple troubleshooting, for example, if the speakers are not working.
• Download and play streaming media
• Download games

In addition to the above task achievement, local teachers and field observers
often note that the children demonstrate improvements in:

• Enrolment, attendance and school examinations, particularly in subjects
  that deal with computing skills
• English vocabulary and usage
• Concentration, attention spans and problem solving
• Working together and self-regulation

The Findings

Groups of 6-13 year olds do not need to be "taught" how to use
computers. They can learn by themselves. Their ability to

do so seems to be independent of their:

• Educational
  background
• Literacy levels in the English language or any other language
• Social or economic level
• Ethnicity and place of origin, i.e., city, town or village
• Gender
• Genetic background
• Geographic location
• Intelligence

Minimally Invasive Education (MIE): How does it Work?

Certain common observations from the experiments described here suggest an
interesting learning process, in a minimally invasive environment where children
self-instruct each other in computer usage:

• Sometimes, one child knows a little about computers already, he shows off
  his skills to others. Sometimes, one child explores randomly in the GUI
  (Graphical User Interface) environment, others watch until an accidental
  discovery is made. For example, when they find that the cursor changes to a
  hand shape at certain places on the screen.
• Several children repeat the discovery for themselves by requesting the
  first child to let them do so.
• While in step 2, one or more children make more accidental or incidental
  discoveries.
• All the children repeat all the discoveries made and, in the process, make
  more discoveries and start to create a vocabulary to describe their
  experience.
• The
  vocabulary encourages them to perceive generalisations ("when you click
  on a hand shaped cursor, it changes to the hourglass shape for a while and a
  new page comes up").
• They memorise entire procedures for doing something, for example, how to
  open a painting program and retrieve a saved picture. They teach each other
  shorter procedures for doing the same thing, whenever one of them finds a
  new, shorter, procedure. They discuss, hold small conferences, make their
  own timetables and research plans. It is important not to underestimate
  these.
• The group divides itself into the "knows" and the "know-nots",
  much as they did into "haves" and "have-nots" in the
  past. However, they realise that a child who knows will part with that
  knowledge in return for friendship and exchange as opposed to ownership of
  physical things where they could use force to get what they did not have.
• A stage is reached when no further discoveries are made and the children
  occupy themselves with practising what they have already learned. At this
  point, intervention is required to introduce a new "seed"
  discovery ("Did you know that computers can play music? Here let me
  play a song for you."). Usually, a spiral of discoveries follow and
  another self instructional cycle begins.

In order to meet the above instructional objectives, it is
important that:

• The computer should be in an outdoor, public, and safe
  location. Children, and often their parents, are apprehensive of enclosed
  spaces such as closed rooms or "clubs". Locating computers
  indoors, even inside a school, is associated with regimentation, control,
  "studying", pedophilia and other negatives that formal schooling
  seems to suggest. Locating a computer in a school playground, on the other
  hand, is ideal.

• Children should use the computer in heterogeneous groups.
  Since the MIE process depends on exploration and discovery, working in
  groups is essential. Collaborative constructivism is the main paradigm of
  MIE. Children teach each other very effectively and are also effective at
  self-regulating the process. That is how over 100 children are able to use
  one computer.

• There should be no adult intervention or supervision.
  Adults should not use the kiosk. All activity should be monitored remotely
  to ensure that the kiosk is being used for the right purposes.

• PC functioning and Internet connectivity should be
  reliable.