Just finished reading one more article in my quest for understanding Alan Kay's Xerox park advances (I should in the future write a post about the quest itself).

I'll summarize here a part of the article, particularly the part regarding education, a little enticement for the whole deal. Sincerely, just go ahead and read the article if you are the least curious about the words here.

So, Alan's words about the principles they used for education in the Xerox park:

These ideas are found in many places and many cultures. We came to them from our own experiences, the Suzuki violin method, O. K. Moore, Piaget, Furth, Bruner, Minsky, Papert, and others.

Here are them, enumerated:

1. Listen to the student.

Since we believe that teaching involves helping a student adapt his knowledge structures to a new situation, we can guarantee ourselves (not to mention the student) an unpleasant journey if we don't try to understand these gossamer schema at the outset.

Many of the current ways that things are done in Smalltalk come directly from listening to the kids. Smalltalk, as an "extensible" system, can easily "be" any kind of tool that we wish. We ourselves have remolded it several times.

2. Never teach anything which has to be unlearned later.

In our experience, humans are very poor at unlearning any kind of skill, whether it be muscular or mental. This principle is well understood by every teacher of music. "Tempting analogies" which later come back to haunt are especially to be avoided.

We teach "straight" Smalltalk, the very same system which adults learn. The very first examples and methods to which the kids are exposed resemble strongly the most sophisticated adult systems.

3. Never pace a student in a way that will require future remediation.

Principle 2. basically says: don't simplify to the point of a lie; Principle 3. is a corollary of this which states: don't put the student into a situation where he will feel dumb and inept because a good enough foundation has not yet been laid. Most kids do not understand the distinctions between skill, structure, and intelligence any better than adults do and are apt to feel stupid rather than unskilled in new situations.

4. Hook on to existing fruitful structures when possible; if unfruitful concepts exist, don't unteach them, rather supply completely fresh orthogonal concepts.

Most kids know about dictionaries and looking up the meaning of a word. The meaning can be an explanation of a passive relationship or a dynamic act. In fact, every idea in mathematics and in programming can be easily explained in dictionary oriented terms alone; this is a fruitful, useful concept, and it makes sense to use it with kids.

Many other "natural language" linguistic structures are ultimately deadly and we avoid them.

Examples are: "nouns", "verbs", "pronouns", inflections, and their counterparts in most programming languages: data structures, functions and control structures, variables, tagging type to names, etc. Instead, we immediately give children a running example which directly exhibits the more fruitful notions of states-in-process communicating-with-messages found in Smalltalk.

5. Do not look over the student's shoulder.

Aside from the obvious reason of avoiding "putting the student under the gun", there is also a great difference between performing and creating. In music, this is known as the difference between improvising and composing (and a greater difference could hardly be found, as any musician will attest). We are much more interested in the design-oriented and planning processes associated with unhurried goal-directed reflection than in the more shallow though flashy effects obtained by virtuoso "thinking on one's feet".

6. Teach and Show Multiple Perspectives of Situations.

A typical problem with fledgling designers of all ages is a strong tendency to commit all of their short term memory to a given perspective of a Situation. If it happens to be an unfruitful view it might be very difficult for them to "bail out" or even tell that it is unproductive. We feel that the Piagetian example of the tilted glass is much more the result of lack of practice in multiple viewing than the result of physiological immaturity.

One of the striking things about design methodology is that "simultaneous" use of a perspective and its dual is remarkably more rewarding than using either separately.

A very global example is the duality of wholes-as-collections-of-parts found in Western science and wholes-as-wholes found in Eastern philosophic thought.

The former has an important dualistic aspect itself: analytic (or top-down) vs. synthetic (or bottom-up); both of these emphasize differences and boundaries: a corpuscular theory. The Eastern philosophy emphasizes samenesses and connection: a field theory. As more complex systems are studied, the apparent differences between the two schools of thought blur in the underlying sameness that characterizes duals

The content of the computer is descriptions of processes; the ability of computers to simulate the details of any descriptive model means that the computer, viewed as cc medium itself, can be all other media if the embedding and viewing methods are sufficiently well provided.

Those principles show a very pragmatic way to teach any skill, not the least programming, mathematics, epistemology and so forth. They are imbued with respect for both the student and the concepts needed for a citizen in the knowledge age to make informed decisions.

You can also see the notion of the computer as a form of media (think speech, printing press, dance, etc) and how that concept is a very powerful idea, that can give rise to a new age of human perception and civilization. Again, read the article to gain a bit of insight on how a group of 25 people 30 years ago created tools and concepts that kept entire fields busy ever since.