some news

Alan Kay alan.kay at squeakland.org
Sat Apr 22 18:21:04 UTC 2006


Hi Daniel --

At 10:02 AM 4/22/2006, Daniel Vainsencher wrote:
>Moving sideways in this discussion - as long as mentoring is more 
>important than content ...

I'd say that if you have good content, the gating factor for success is 
definitely that quality of the mentoring for most children.

>, the number of kids that can be "infected" at any given time is bounded 
>by the number of people able to do the mentoring. I expect mentors are 
>even rarer than computers where the 100$ laptop is headed.

Yep, even in US, Europe, Japan, and generally worse in 3rd world.


>Haven't there been any serious attempts to make systems in which the 
>content itself takes the user on a reasonably long and useful ride even 
>with no mentoring?

Yes, and it works for some children, especially the 5% who are wired enough 
to "almost be there" and for those who have learned how to learn via 
reading (I was one of those).

A very large number of children don't read well enough or have enough 
interest to follow directions. Most also have very few options when they 
get stuck. For most children, mentoring of some kind (including from the 
system if it can) is the bridge over the gaps.

What Jerry Bruner called "scaffolding" is a little different from child to 
child, but it is key.


>I had about 3 years of fun with Basic and about 500 pages of exercise 
>booklets, and nothing more. I think two important elements to that success 
>were that the Commodore64 had zero extraneous interface, and that the 
>booklets started from explaining the keyboard.

And, most likely, that you were in that 5%! Adele and I realized early on 
that the real key was to find out what to do for the next 85%, and this is 
where actual pedagogy and educational environments (and mentoring) really 
matter.


>Why can't the booklet be part of the environment, and improved until 
>mentors can be banned from the room with no/little ill results?

Check out the idea of "Active Essays" that Ted Kaehler and I starting doing 
about 12 years ago. These bridge some of the gap for readers. But are not 
very useful for non-readers.

The UI is part of the mentoring environment (this is why I got interested 
in naive relatively easy to use and explorable UIs in the 70s -- they were 
initially for children).

If you don't have much of an AI behind the scenes, then a really great UNDO 
makes a huge difference. I think that was one of the main PARC 
contributions to UIs. (Squeak does not have a great UNDO ...).

But, aside from having huge mechanisms that can anticipate and deal with 
user errors in a graceful way (c.f. Anderson's work at CMU with tutors of 
various kinds), not much general mechanism exists. I have thought several 
times about getting Anderson to do a massively gentle tutor for just the 
first 30 minutes of the Etoys experience (and I still think this would be a 
good idea for the $100 laptop).

Among other things, the Nebraska facilities in Squeak Etoys (and the more 
comprehensive collab facilities in Croquet) are there partly for the 
purpose of allowing children who have some experience somewhere in the 
world to help children with less experience in other parts of the world. In 
McLuhan's global village, this is the analogy to the one room schoolhouse. 
Both Seymour and we have experimented with this over the years and are 
convinced that it is a part of the solution of the puzzle.

Cheers,

Alan


>Daniel
>
>Alan Kay wrote:
>>I agree that content is really important, and even more so is mentoring. 
>>The language is less so providing it doesn't develop limited ideas (like 
>>BASIC did).
>





More information about the Squeak-dev mailing list