[etoys-dev] (SQ-749) and Kathleen's question on "What do you mean by Artifacts?"

Alan Kay alan.nemo at yahoo.com
Mon Aug 23 07:57:30 EDT 2010

Hi Steve,

Thanks for the thoughtful note!

By "when easy and when hard?" I meant parts of the pedagogical and learning 
processes that have been tested out ahead of time. For example, Ted Williams 
(one of the greatest hitters of the last 70 years) determined through experiment 
that using a batting tee (which he invented) would really help one very 
important aspect of learning to hit. On the other hand, it has also been 
determined by experiment that putting frets on a violin doesn't help, and is 
even retrograde. There are many such questions in mathematics, science, and 
especially computing. In writing, it is not thought to be a good thing to use a 
"library of already written sentences and paragraphs" to aid composing a piece. 
But it is done all the time in computing (and sometimes this is helpful for some 
of the reasons you list, and sometimes it is a very bad idea because it makes it 
easy to ignore building some of the important fluencies).

I still don't recognize the quote .... but if you send the Elluminate video to 
me, we can certainly do a transcript (there is a very speedy and reasonably 
inexpensive transcription company here in LA).

Criticizing teachers and adults accurately is not bashing. And the real problems 
have to be faced. I was very lucky to have learned to read fluently and rapidly 
before I entered school, and this gave me a way -- via home and town libraries 
-- to bypass "the barriers of adults in learning" pretty effectively except for 
too many unpleasant personal interactions. That being said, I also had several 
exemplary teachers who made a big difference in my life (and you can't 
substitute books for these). But the ratios of bad vs good were horrendous, and 
books can be read and discarded rapidly to find the good ones. The wide range of 
personality and motivational types we find in children suggests that "better 
adults" are really critical for many of them -- not all of them will be as 
driven as I was to get around the barriers.

The "older kids" path is a good one, and can help a lot, and much more should be 
done with it.

The other thing to ponder is that it is philosophically possible to make a 
technology that lies somewhere between what books already do -- bring us 
important parts of genius teachers and minds in a form that can be replicated 
and distributed by the millions -- and what great human teachers can do. Exactly 
how far this can go and what kinds of people it can work with is yet to be found 
out. For example, could we imagine "a book that can help kids learn how to read 
it"? What if we took (say) Dr Seuss books and "did something" that could help a 
child to read them in a variety of ways? (And by the way, here is a perfect 
example of "when easy, when hard?", because an overshoot of giving too much help 
could remove the motivation of the children to learn how to read themselves.)

To pick neutral ground, a lot of really valuable help in learning music could be 
given by "superbooks" -- and this is partly (a) because there is a lot known 
about how to teach various aspects of music (especially learning how to play an 
instrument), and (b) because this is one area in which "good listening" can 
already be done by computer programs. ("Good watching" is not so easy to do at 
this point.) And (c) because the computer doesn't have to do all the listening 
-- it can help the learner listen as well in a variety of ways. This is a key 
idea in several language learning environments available now.

Some subjects will be quite challenging to do "good watching" in for some time 
to come. But quite a bit of mathematics and especially that which overlaps with 
special kinds of programming have the possibility of doing good watching, so we 
can expect to start seeing much better environments for children over the next 
10 years or so.

Re Maria Montessori -- she struggled to design her environment as well, but she 
stuck with it (and was a special kind of genius), so she got many good results. 
She is definitely a patron saint of ours ....

Re Galilean gravity ... For now, I suggest the much easier route of making "jpeg 
movies" (they are just a folder of images whose names sort in order). The Etoys 
movie player has an option for playing such jpeg files.

One of the projects for etoys that is illuminating is to make a simple movie 
player using animated images. (This is a perfect example of Etoys needing to 
have a way to package up objects (it actually does, but it is not very 
convenient -- the "world menu" has an option for saving morphs, and all the 
connections between them will be preserved if the objects are put in a Playfield 

That said, I completely agree that "objects should be easy to package, reuse and 
to share".

I think I'm not understanding your graphing example. It is "pretty easy" to 
position text objects containing numbers (you can adjust the centers of any 
object and the x and y coordinates can have some arithmetic done to them with 
results put in the numeric value). Please give me more info here.

I don't think "just everything" should have to be made from scratch in order to 
learn -- but I also think that one of the reasons that so many adults don't get 
fluent at this stuff is that they don't make enough things from scratch. 
(Analogies to drawing and painting, sports, music, etc. should be drawn here.) 
And, sure, they are already busy and there isn't a lot of spare time -- so the 
bootstrapping process is difficult when the adults start late, after they are 
already busy. There are chicken and egg problems also. The more fluency that is 
attained also changes perceptions and realities about easy and hard.

To go to science, most scientists learn most of their knowledge from reading, 
but it is the "real science" they do when they are starting out that changes 
what they are doing later from "believing in a new catechism" to "actually 
thinking as a scientist about scientific relationships". Similarly, there is a 
certain amount of "To know the world, one must construct it" (Pavese) that is 
absolutely critical in learning mathematics, where part of the main point is to 
chain together inferences and to understand how claims are preserved via that 

Re: the next Etoy-like environment. The current plan is to work on a "science 
and systems" curriculum for ca 8th grade and to make a computer environment that 
can really make a difference for this curriculum. This will not be an extension 
of Etoys, but will be more comprehensive in many areas, and will have new 
abilities in others. It should wind up being better than Etoys for the 5th 
graders that Etoys was originally aimed at, and it should fix many of the 
problems that we all agree are there in the current version.

Please do follow through with your plan of asking for 3 (or 20) things -- 
thoughtful comments like yours are like gold for us.



From: Steve Thomas <sthomas1 at gosargon.com>
To: Alan Kay <alan.nemo at yahoo.com>; Bert Freudenberg <bert at freudenbergs.de>; 
Kathleen Harness <kharness at illinois.edu>
Cc: etoys-dev <etoys-dev at squeakland.org>; squeakland <squeakland at squeakland.org>
Sent: Sun, August 22, 2010 8:37:21 PM
Subject: Re: [etoys-dev] (SQ-749) and Kathleen's question on "What do you mean 
by Artifacts?"

Hi Alan,

On Thu, Aug 12, 2010 at 12:54 PM, Alan Kay <alan.nemo at yahoo.com> wrote:

A good slogan for teaching, pedagogy, and curriculum design is "When should it 
be easy, and when should it be hard?". 


This, for me at least, is hard to figure out ahead of time when designing a 
lesson. I am much better when I interact with the kids and can improvise on the 
The notion here is that for education to be transformative, you wind up as a 
different "better" person than you were, and this means that certain 
difficulties were very important to your learning -- the stuff that was "easy" 
you could already do (and some of the easy stuff is not what you want to aim 
at). Building knowledge and skills to get fluent at handling difficulties (and 
in some case rendering them non-difficult) is a key for much important learning. 


>On the other hand, gratiutous non-productive difficulties are to be avoided 
>because they generally both distract  and occupy " thinking chunks" that one 
>needs for the important stuff.
>(I don't recognize my quote that was paraphrased)
>It comes from a video I found in the Bell Labs library around 84-86 (not sure of 
>video date). You were talking about User Interface design and my biological 
>memory recalls you refering to using a mouse and saying "If I can do it in one 
>click, I'll do it.  If it takes two clicks, I might do it. If it takes three or 
>more clicks, I probably won't do it". I have used and attributed it to you a 
>number of times since. If I am in error please let me know and I will self 

I really could use Gorden Bell's E-Memory. Hopefully you are heeding his advice 
and collecting your E-memory so that we may search and study it. One way to do 
this is anytime you give an interview, presentation etc where it is recorded, 
request a copy of the recording and all rights to distribute as you see fit. 
 Also be sure they video your presentation and not just you so we can see what 
you are referring to.  It would be nice if we had text versions of those 
interviews. I am working on transcribing the interview you gave at Natural 
math and its time consuming. I have a friend who works on Google speech 
recognition software and would try and send him a copy, but Ellumintate doesn't 
seem to have a way to extract the audio. 
Adults tend to be the biggest problems when trying to help children learn 
things. It's the adults who generally don't want to do the work and don't see 
many things as fun. Kids (and people in general) can spend a lot of time 
focussed and doing when they are having fun.
 If not Adults, who? If the goal is transformation I see Adults as part of the 
solution.  So does labeling them as part of the problem (which I am NOT 
disagreeing with) help or hinder transformation?

I ask this because I see a lot of teacher bashing which I believe hinders our 
cause much more than helps it.

Another choice besides adults is older kids, I have seen this succesfully used 
in Scouting and in a public speaking conference my kids attended this past 
weekend. They tend to listen to older kids more and the older kids realize the 
importance of fun. The key here, as with adults, is guidance and training.
Constructivism (one of many such terms I don't use because they have lost their 
meanings) doesn't mean discovery from scratch (this is a huge confusion many 
people have), but does mean "understand and clarifying by making a careful 
descriptive model". 

This can be done with English and writing (it is what descriptive, expositional, 
and argumentative writing are supposed to be about). It can be done with 
mathematics. It can be done in many cases with physical construction materials. 
And a lot can be done in terms of computer programs.
>Thank you I really hadn't thought of this before and it is an excellent point 
>(ie: something I can use with my kids).

I will look for ways my kids can learn this.
I like the  Montessori curriculum approach of making a carefully designed 
environment for the chilldren that allows choice on their part and allows 
limited the degrees of freedom on the educator's part.

The problem I struggle with is how to design the environment and what games, 
playthinks, problems, etc to use.
One of the best projects we've ever designed is the Galilean Gravity one -- and 
it illustrates what you have to do with guidance on the one hand and space to 
play on the other to enormously raise the probability that most children will be 
able to see and understand what is going on without having to give them the 
"answer" to memorize.

I love that project and I find it extremely frustrating. Frustrating because I 
really want to do it with my kids, but it is really really hard to do in Etoys. 
 The hard part being importing the video. The reason for this I believe is that 
the video player only supports MPEG.The idea that kids take their own videos and 
then analyze them is an excellent approach. Avigail Snir had a wonderful example 
of using video at Squeakfest.  But as I understand it, even she had a lot of 
problems trying to convert the video into MPEG and even more frustrating is she 
is having trouble sharing it with the world so other teachers could see it and 
be inspired.

One of the keys in the early Montessori schools was the intense comprehensive 
training of the Montessori teachers -- and the lack of the equivalent of this in 
most of today's schools is a huge problem.
I have also been thinking about using scripts (the theatrical kind) and cue's to 
look for with scripted responses as a way of teaching. Perhaps Actors would be 
good teachers hmmmm need to flesh out the idea more. 
The scripts in Etoys are independent of the visible appearance of the object 
they are attached to (the objects' "costumes" can be changed at will -- and this 
is how "frame animation" is done in Etoys). 
I'm not quite  understanding what it is that you would like for teachers beyond 
a repository of projects with extensive notes about how they were made and how 
to make them.
A repository of projects is good, but I would add a repository of objects, sets 
of objects and scripts. Ie: I think unit of "things you can share/store/easily 
re-use" is too large. I write this then ask myself, why do I feel this way?  I 
think it goes back to the point you verbalized much better than me "It gets 
annoying after the first few times".  I may be thinking of what you refer to 
later in the email as "packaging up" a solution.  

I would like to be able to re-use and share those solutions (especially since 
Etoys lets me look inside and modify them). The ability to look inside and 
modify them is why I would prefer "artifacts" (as discussed in previous emails 
and built by Hilaire as part of iStoa) be created with scripting tiles as 
opposed to doing so in squeak.  This would allow teacher to build a set of 
virtual manipulatives that others could modify or just look inside and figure 
out how they work. One response I received was that this would not scale, but 
frankly I don't understand why not.

"Packaging a solution" would also let you develop a curriculum where kids built 
their own tools (ex: a graphing tool) and re-use them in other projects. The one 
tricky part of the graphing tool is labeling the number lines (trying to align 
the text is hard, at least I have not found a simple solution) so perhaps a 
number line object would be needed.
If "learning by making" is a good idea, then shouldn't teachers learn new ideas 
about Etoys by making them (but with lots of guidance)?

Okay my turn to not quite understand: Yes I agree "learning by making" is a good 
idea. But should teachers and learners have to make everything from first 
principles?  Who has the time?

Would teachers not benefit from using "great literture" created by people who 
were PUFx's (ie: had a Profound Understanding of Fundamental <subject matter>)?
I have a friend at work who while at Stanford and did very well in Organic 
Chemistry, he thought the reason he did so well was he had 7 or 8 core concepts 
from which he could derive the rest.  He decided to try and teach those concepts 
and when and how to use them to some of the other students to see if they could 
use them and "transform".  It worked well and I think this idea has some 
merit. Unfortunately he had a head injury and can't remember them. My challenge 
is to find sets of core ideas and how to use them, then figure out how to 
present them in an age appropriate manner.

I also think you made a good point in the Natural Math interview when you 
mentioned Seymour Papert's comment that it doesn't much matter what mathematics 
we teach them as long as they are learning to reason like a mathematician. 

On the other hand, there are any number of things in the Etoys design itself 
that could be vastly improved to help both adult and child learning, and also 
for them to make better extensions.
Yes, the question I am struggling with are which are the important ones worth 
doing first. To think about that I am considering imposing upon myself something 
similar to the Warren Buffet advise to "give myself a ticket with only twenty 
slots in it so that you had twenty punches - representing all the investments I 
could make in a lifetime". I think instead I will try and limit myself to three 
requests for the rest of the year. Those who know me know how hard it will be to 
impose that constraint. But one way would be allow myself to discuss many 
possible improvement (which helps me think about them) and then only pick three 
to request.
For example, I do all my talk presentations using Etoys and I write scripts to 
sequence "builds" of additive visual material to the slide (Powerpoint has a 
feature for doing this that is more convenient for some goals and simply won't 
allow others). It is very instructive to do this by hand a few times, and then 
gets annoying.
Etoys does not have a good extension mechanism for "packaging up" a solution to 
"slide builds" that can then be used as a feature. This is a real  sin against 
our own precepts. The lack of it is due to EToys being thought of as temporary 
and of limited scope at Disney. It is terrible that we don't have it now. 

I think Ricardo's GSoC Morph I/O may help solve at least part of the problem. 
 You can save an object (including a playfield and its embedded objects) in a 
file and import it later. I have used it a number of times and find it very 

Why don't we do it now? Because we've been trying to move on to the next design 
since before OLPC came along. And so forth.

I look forward to seeing the next design (I have only caught glimpses from what 
I have read on the VPRI site and reading into emails and discussions). Is there 
a time frame when we can expect to see it?
All that said, I think the most important thing to work on now is great content 
(projects, lesson plans, screencasts, etc.)

Many Cheers for all you have done and continue to do,

From: Steve Thomas <sthomas1 at gosargon.com>
>To: Bert Freudenberg <bert at freudenbergs.de>; Kathleen Harness 
><kharness at illinois.edu>
>Cc: etoys-dev dev <etoys-dev at squeakland.org>; squeakland 
><squeakland at squeakland.org>; Alan Kay <alan.nemo at yahoo.com>
>Sent: Thu,  August 12, 2010 8:11:46 AM
>Subject: Re: [etoys-dev] (SQ-749) and Kathleen's question on "What do you mean 
>by Artifacts?"
>I have been reading Alan Kay's Thoughts About Teaching Science and Mathematics 
>To Young Children:
>I think one of the trickiest issues in this kind of design is an analogy to the 
>learning of science itself, and that is "how much should the learners/users have 
>to do by themselves vs. how much should the curriculum/system do for them?" Most 
>computer users have been mostly exposed to "productivity tools" in which as many 
>things as possible have been done for them. The kinds of educational 
>environments we are talking about here are at their best when the learner does 
>the important parts by themselves, and any black or translucent boxes serve only 
>on the side and not at the center of the learning. What is the center and what 
>is the side will shift as the learning progresses, and this has to be 

By exposing everything in Etoys as "First Principles" (which in this particular 
case I understand to mean, that we have a minimal set of scripting tiles and 
objects from which everything can be built) we avoid the "productivity tools" 
issue because everything is exposed.  It is also a beautiful, elegant and 
exposes a powerful idea.

The challenge in a system where everything is done from "First Principles" is 
that when you are designing an"educational environment"  "lesson", or "Artifact" 
( better terms might be "playthink" and/or "tool to think with"), it can take a 
lot of work to build those preferably translucent boxes.  And to paraphrase 
another Alan Kay quote on user interface design: "If it takes one step I'll do 
it, If it takes two steps I might do it, if it takes three or more steps forget 
about it!"

No, I am not arguing to make things easy for everyone, we need find ways to get 
kids to have "hard fun." Hard work and ragging a problem are good habits.  I 
also strongly believe that giving kids a "blank canvas" and a great set of 
brushes and paints is an excellent and preffered method, but not the only one we 
should use.

I am arguing (and struggling) with is how in a "First Principle" system like 
Etoys, we can find ways to make it easier for teachers/designers (ie: make them 
more productive).  I fear I see in some folks (none on this email list of 
course) a tendency towards what I initially saw (and fell into myself) as the 
constructivism trap. Where I encountered people who thought kids should 
construct all knowledge themselves from Scratch (pun intended ;).  As I recall 
Alan (and others pointing out) we can't expect kids to do that, they will repeat 
the same mistakes people did over thousands of years. My initial thoughts are a 
repository of Artifacts that teachers can use along with a set of scripts (the 
problem with the set of scripts idea is that the scripts in Etoys are not 
decoupled from the ?morphs? (not sure of the correct term here, but basically 
the pixels visually representing the object). Bert's idea that we have a Player  
Variable and the scripts that operate on it is a good one, but I think there may 
be some bugs there, need to test more.

Now I will more directly address Kathleen's question: "What do you mean by 
I will switch from "Artifacts" to the term "Playthinks" (which I encountered in 
the "The Big Book of Brain Games" by Ivan Moscovich).
One of the best and simplest "Playthinks" for teaching I ever encountered was 
Robert B. Davis' classroom warm-up (which I showed at Squeakfest and have wrote 
about here.)  Basically it involves drawing on the board a 4 x 4 grid
. . . . 
>. . . .
>. . . .
>. . . .
Then having kids pick two numbers and using those two numbers as X and Y 
counting from 0 at the origin point in the lower left and then If they land on 
the board marking an X or O until one team wins.  Some of the keys to this 
"Playthink" are:
	* you let kids puzzle it out for themselves, they figure out the rules, you 
don't tell them
	* it contains a powerful idea
	* it can be easily extended to other concepts (negative numbers, a number is 
all the ways you can name it, is this game fair ...)
	* its fun 
Other examples of "Playthinks" would be cuisenaire rods, pentagrams, area 
blocks, other good "virtual manipulatives" and my feeble attempts Circle 
Explorer and Pattern Blocks and Tools.

Bert, your comments on SQ-749 sparked my writing this, I will address it more 
specifically in a separate email after some more thought.

Including Alan so he can correct any misinterpretations and hopefully comment

FYI: A lot of other excelent writings from VPRI are here, most are Computer 
Science related but a number deal with educational issues and Etoys.

On Thu, Aug 12, 2010 at 5:02 AM, Bert Freudenberg <bert at freudenbergs.de> wrote:

>On 12.08.2010, at 10:32, Steve Thomas wrote:
>> On Thu, Aug 12, 2010 at 3:32 AM, Bert Freudenberg <bert at freudenbergs.de> 
>>> This would likely be simple to implement, but could break existing projects. 
>>>E.g. the morphing stuff in the showcase. Maybe you could take a look?
>> Checked "Morphing" by Kazuhiro Abe and that should be fine. Each Polygon in the 
>>holder has the same number of vertices and the script simple changes the 
>>positions of the vertices one at a time.
>> "The Walkers" and associated remixes by P.A. Dreyfus uses a special category 
>>"morphing" which would be great to get into Etoys (although I would suggest the 
>>addition of some way to show/manage the frames) to make the invisible more 
>>visible and to make it easier to create these kind of animations. Anyway, 
>>whether this would be a problem or not depends on how it is implemented.  If he 
>>stores complete information about the polygon in each frame, I see no problems. 
>>If he only stores differences and adds/removes/repositions each vertex that MAY 
>>cause a problem.
>> Anyway if it is a simple change and you can make it, I think I can easily test 
>>the change by opening the project, then file-in the changes and see if anything 
>>breaks. Or you could also ask P.A. Dreyfus (master of polygon's and connectors) 
>>what he thinks as he knows and can check the changes against his implementation.
>> Stephen
>Thinking about this more I do not like the proposal. It would makes the system 
>less predictable.
>Having the new vertex remain at the same position is the only sensible choice. 
>It matches the "copy" behavior of regular objects, which also appear in the same 
>position. It would not scale anyway - see this image where I only inserted 4 
>vertices. The position quickly converges to the next vertex position.
>Also, I'd argue that "add a vertex at beginning" and "add a vertex at end" tiles 
>are not needed in the first place. To be useful they would, as you noticed, have 
>to be "set cursor to beginning and insert vertex" and "set cursor to end and 
>insert vertex", because otherwise one cannot assign their position immediately. 
>But that makes them perform two operations that are available separately. There 
>is no good reason to coalesce those steps into one.
>In any case, inserting a vertex should not change the cursor. If you want a 
>cursor change to occur, insert a tile.
>So my counter-proposal is: remove the "add vertex at beginning" and "add vertex 
>at end" tiles. (to not break existing projects, the tiles would only be hidden)
>- Bert -
>etoys-dev mailing list
>etoys-dev at squeakland.org


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.squeakland.org/pipermail/etoys-dev/attachments/20100823/612205eb/attachment-0001.html>

More information about the etoys-dev mailing list