Squeak ideas for a classroom/clubhouse

[Alan Kay]

Hi Jahanzeb --

I'm going to write two replies to your very nice and interesting 
letter. This is the first, and I'll try to follow up with a more 
thoughtful one in a few days.

At 7:20 AM -0800 1/11/03, Jahanzeb Sherwani wrote:
>Hello all,
>
>I'm a research associate at LUMS University, Pakistan, and am working on a
>research project here where we're trying to use low threshold/high ceiling
>software environments to enable schoolkids to create and learn, and have
>fun while doing so. I'm interested in giving the fifteen kids (aged 7 thru
>14) the flexibility to create whatever it is they want to create: graphics,
>music, games, stories, building their houses or other aspects of reality --
>such that their learning is made personally meaningful (along the lines of
>Papert's Constructionism) in ways that are rarely found in education in the
>Third World. However, we'd also like to give them challenges that will
>bring out the most learning (such as learning about feedback through the
>cars, as shown in the elementary school gallery by BJ). Our main reason for
>choosing to do such a project is to enable the kids to think outside the
>hold of a curriculum that holds little to no relevance to their daily lives
>or the world around them, and to help them do stuff that's educational,
>fun, and personally relevant.
>
>I was initially using Alice2 as the major environment to work in, although
>I'm running into quite a few problems (namely, it's not running on the
>school's PCs for some reason!) so I've been looking at other options, and
>discovered Squeak (which many at the MIT Media Lab recommended when I
>visited a month ago), and I believe it perfectly fits the bill.
>
>I managed to go over parts of the mailing list archives before asking these
>questions, so I hope you'll forgive me if they were already addressed
>before. My questions are:
>
>1) How does one act as a facilitator in such an open setting with Squeak,
>so as to allow diverse views of what each wants to do, but still make sure
>that there is some learning (and not just air guitaring) going on? There is
>so much you can do with Squeak (no ceiling), but how does one try to nudge
>it along lines that will lead to good learning -- or is that a
>contradiction in terms?

What we try to do is inspired by Montessori: to come up with projects 
that the kids absolutely treat as toys and play, that also have (we 
think) beneficial cognitive side effects.  So all the stuff about 
cars and driving, the car races, etc., teach about vectors, velocity, 
accelleration, the idea of "random", the idea of feedback (to make a 
robot car that can stay on the road), etc. Our experience over the 
last 3 years has been that virtually every child in a classroom gets 
really interested in this stuff for their own reasons. Recent 
experiences in Japan indicate that these projects and the process are 
pretty independent of simple cultural distinctions.

Another area that children love is nature, especially regarding 
animals. And it is a source of delight to them to find out that they 
can make a new costume for their car and turn it into a fish or 
horse, etc.

Last year we tried a more ambitious project in science. This involved 
having the children (10 years old) first learn about velocity and 
accelleration using their cars. This is a very nice project all by 
itself. Then we had them do some measuring (for example, a bike tire 
circumference) using different tools. This helped them understand 
that measuring is likely to not produce the exact same numbers, but 
is likely to produce numbers similar in magnitude.

Then we showed them various objects (two shotputs of different 
weights, a croquet ball, a foam ball, some apples, etc.) and got them 
to speculate as to which would fall faster or slower. Then we took 
them outside and dropped the objects from the roof of the school. We 
took videos of these drops. The videos were imported into Squeak and 
the kids could look at every 5th frame and measure what a dropped 
object was doing. They could see that the pattern was the same one 
they had seen when they were doing accelleration with their cars. 
This led them to write similar scripts to accellerate a painted 
object to match the movie. (We have a nice video clip of this 
process.)

Thus they were able to experience a phenomenon of the real world, 
measure it, put a model to it, and make a mathematical simulation of 
it. (Most American college students are not at all successful at 
learning this using standard methods.)

Once they had a script that would move objects as gravity moves them, 
they had a new tool and toy to make gravity games, etc.

>
>2) How does one introduce the medium as something that is infinitelly
>malleable, and that it is ok to add/change something if you don't like it?

Most children have a big revelation about this in their first few 
hours of doing stuff. We've noticed it happening many times when they 
put a new costume to their car object and realize that they can make 
anything and make it do anything.

>For instance, the lack of a 'move sideways' tile (like the 'forward by')
>tile means that kids will have to start off controlling their
>game-characters with a move forward/backward, turn left/right instruction
>set, and so can't start off by making a Pac-Man type game (which needs to
>move left/right, and not turn). Should I create a 'move sideways' tile
>beforehand

I would suggest not.

>, or try to help them make it themselves as they require it?

There are two approaches here and both are worthwhile in the end. The 
first is that using "object forward by" and "object turn by 90" will 
do what you want. The second is that there are x and y location 
properties in the viewer of all objects. It is very worthwhile for 
the children to see that:
       object's x increase by 10
will move the object 10 pixels to the right, and that this is exactly 
equivalent to
       object forward by 10
if the object is pointed to the right.

Similarly,
      object turn 5
is exactly equivalent to
     object's heading increase by 5

>
>3) The main reason I pushed to have classes of diverse ages was that the
>young ones will be able to learn from what the elder kids are doing, and
>will also get a sense of self-respect by working on the same environment as
>elder kids. Should we be giving different problems to the younger ones to
>solve, or not? Could anyone on the mailing list (who has experience with
>such age ranges in such classes) tell me about their own experiences, if
>possible?

Teach the older ones a few days before the younger ones. You can 
start with the same set of projects, but the older children can go 
quite a bit further and faster. So it's a good idea to have more 
project ideas for them.

>
>4) The social relevance of education is something that is touched upon most
>by Paulo Freire, who said that imported curricula aren't adequate because
>they lack relevance to the sociocultural environment, particularly in
>developing countries.

This is an interesting claim. I think it is true at one level, but it 
rapidly misses the point once education starts to happen (and this is 
the great difference between "education" and "training"). Perhaps a 
milder view is that in any kind of user interface experience, the 
designer has to start in the world that the endusers live in.
      However, the learning of powerful ideas is not just a new tool 
that one wears on one's belt, but an actual change in how the world 
(especially of ideas) is perceived. It's a change of perspective as 
well as one of knowledge. A child who learns science starts to become 
part of a different cultural environment, and this is why scientists 
quite resemble each other and can easily communicate with each other 
all over the world regardless of their initial background.
      This isn't the same thing as impressing "Western Civ" on other 
cultures -- it's partly an accident that science was invented in 
Northern Europe (it could just have easily have happened one or two 
thousand years earlier in the Mediterranean or in China or Japan).

The simple bottom line here is that young children especially are 
interested in things they can *do*. So you will have no problems. 
Freire was talking much more about trying to educate adults who had 
grown up in traditional cultures (and here, I think, he was most 
right).

>  Does anyone have any experience in addressing these
>concerns best through the use of environments such as Squeak?

I think Mitchel Resnick of the Media Lab has had more experience than 
we have, with his various LOGO in schools projects in Latin America, 
and his more recent computer clubhouses around the world.

>
>Our twice-weekly sessions with the kids in their school labs begin this
>Tuesday, so whatever you could tell me before then would be extremely helpful.

I hope this helps. Also Kim Rose has had the many experiences 
teaching Squeak etoys to children and teachers around the world. She 
can provide quite a bit of guidance as well.

Best wishes,

Alan

>
>Jahanzeb Sherwani
>
>ps I agree that Open Croquet looks quite amazing (although quite slow on my
>P3 which was to be expected)! Are there any mailing lists for it yet?


--