Communications of the ACM

---

Ken Kahn

October 27, 2019 11:33

Yes, but there is something special about programming. As Alan Kay wrote in Scientific American in 1984: "It is not a tool, although it can act like many tools. It is the first metamedium, and as such it has degrees of freedom for representation and expression never before encountered and as yet barely investigated."

Fortnite, TikTok, Word, etc. are not metamedia.

Programming can be an especially fertile ground upon which to learn to think and learn.

---

Mark Guzdial

October 28, 2019 10:36

Hi Ken,

We completely agree that programming is powerful. Our first vignette is about the special power of programming. One can think well without programming. We do encourage programming by children, but (speaking for me) the tools, languages, and environments that we have for children's programming could be made better. What would programming on top of Fortnite and TikTok look like? How would we re-invent programming for the goal of "expanded thinking" rather than software development? How would we re-invent programming differently for the goal of expanded thinking in different domains?

---

Alan Kay

October 30, 2019 03:04

Thanks Ken

I think there is something really special about programming also. However, I think we can agree that there is abundant evidence today -- from looking at the large numbers of programmers around -- that just learning to program conveys no automatic special insights (perhaps almost the opposite?)

On the other hand, I know that we do agree that programming of certain kinds used in special ways can convey revelatory insights that can qualitatively lift our levels and kinds of thinking. That was certainly what our great inspiration -- Seymour Papert -- was all about, and he was quite wonderful at being able to find ways to get the children thinking about "the stuff behind the stuff" and not falling into the trap of thinking mechanically.

"User interface" of all kinds -- including many kinds of communication -- has to start with where people are. A really good experience will help lift people from where they are into higher insights and thinking (most UIs today fail miserably at this, and it's a real shame that some of the computer artifacts that children like to play with -- like games -- almost completely miss the opportunity to help children as well as enthrall them).

A simpler parallel is the extremely important longer standing question of how to teach children to read and write. We certainly want to start out with material and activities that are intensely personal, even as we endeavor to help children widen out to deal with ideas that are not so close to home. I don't think schools do a great job of this, but I also think they are doing a much better job than is seen in their approach to computing and programming.

The biggest problem when only the "flat tire" or "half full glass" is in view is that it eventually becomes taken for "normal" and "reality", and this really hurts helping children become aware of the much larger possibilities. If schools were better, we would not have to rely so much on fruitful serendipitous experiences outside of school. As it stands right now, I think getting the computing environment around children to be much much better -- which is very possible to do, and doesn't need school politics to accomplish -- is a path all computerists should rally around to make happen.

---

Ken Kahn

October 31, 2019 12:03

I agree with the comments and much of the article. And I agree that the exploration of adding programming on top of Fortnite and TikTok is worthwhile. Where I think we disagree is about emphasis or focus. In Seymour Papert's introduction to his book Mindstorms he writes about how a powerful programming language can be used for thousands (millions?) of different things. Among those things it is likely that a child will be passionate about a few of them. Yes, many children are passionate about Fortnite (and TikTok?) and some of them would be passionate about programming in Fortnite.

There are have been many successful attempts to add programming to games: Rocky's Boots (1982), Robot Odyssey (1984), RoboSport (1991), Minecraft (multiple extensions), and probably many more. But these efforts excite a small fraction relative to those who are excited about using general-purpose programming systems such as Logo, Scratch, Squeak, ToonTalk, or Snap! for their own projects.

I think our efforts should be focussed on supporting children in pursuing large and deep projects that they are passionate about. Fortnite and TikTok may be distractions from that goal.

---

Alan Kay

November 02, 2019 05:08

Hi Ken

I very much agree.

I will confess that for a variety of reasons I detest many of the apps around, especially those that successfully work as new legal drugs by exploiting our genetically built in wants via oversupplying what used to be safely scarce.

I also thought that Rocky's Boots was brilliant (it didn't quite teach programming, but what lies behind it), and the conception of Robot Odyssey was wonderful (even if the realization of it was too awkward -- someone (like you) should take another shot at this idea). The attempts to tack programming onto Minecraft have been clumsy, but perceived need has surmounted the awkwardness.

Here's one crux that needs to be pondered and resolved vis a vis schooling and society: it's one thing to learn something because it's fun -- a great way to do it for sure -- and another to be helped to get really fluent and really deep at it for larger societal reasons (my reading and writing example above).

I don't think most people -- especially children -- are learning programming deeply enough and in powerful ways that bring forth the profound kinds of thinking -- especially systems thinking -- that citizens of societies need today. Society's official bastion for improving its body politic has always been the schools, and these have fallen disastrously below threshold on any number of fronts.

I don't think our world can survive with just pop conceptions of ideas. I think this is dangerously true regarding pop conceptions of central subjects such as math, science engineering, history, and ... computing.

---

Robert Gotwals

November 03, 2019 03:15

So, in all this excitement about computational thinking, whatever happened to a discussion of computational SCIENCE, or has that been subsumed into computational thinking? At my high school, we have 11 courses in computational science, including new courses in digital humanities and data science, along with computational bio, chem, physics, medicine, nanotech, and a couple of research courses. The goal of ALL these courses? Teach bio, chem, physics from a computational approach. Computational science is a well-established discipline, but it doesn't seem to be appearing much in any of these discussions.

---

Mark Guzdial

November 03, 2019 04:20

Ken, thank you for your comments! I started digging into them, and decided to write a blog post in response: https://cacm.acm.org/blogs/blog-cacm/240605-the-size-of-computing-education-today-by-the-numbers/fulltext

---

Peter Skillen

November 13, 2019 08:14

Thank you all for this insightful article. :-) There are many thoughts I have in relation to this, but the one that immediately jumped out for me is in relation to this point:

""[K-12] Students using word processors for writing generally produce longer, higher-quality
writing than students using pencil or pen and paper." The computational tool plays a role in
students ability to write. We might say that using professional writing tools leads to
performance thats more like a professional writer. Its honest use of the real thing."

This takes me back to my old friend Gavriel Salomon's thoughts about 'effects with' vs 'effects of' computing. It may be that the student produces better writing 'with' the computerbut, is the student a better writer after having used the computer? Has there been a cognitive residueand 'effect of' having used a wordprocessor?

I'm not sure. I think if students perhaps used the 'outliner feature' of wordprocessing, there might be some cognitive transfer to 'whole' and 'parts', 'forest' and 'trees'. The idea of mind-sized bites (thank you, Seymour) might be acquired more easily.

This cognitive residue is indeed partly possible by the computing tools and is partly the result of the classroom culture in which the tools are used. Teachers can help to build the transfer for students. I use use a wee technique I call 'metaphoria' to help young students to think about their programming. I would provide startersalthough they are encouraged to make their own)!

"Programming in Logo is like"...
'...playing tennis. First, I take a turn, then the computer takes a turn.'
'understanding space. The Earth is like a subprocedure inside the solar systemwhich is like its superprocedure. And the solar system is like a subprocedure inside the galaxywhich is like its superprocedure.'

These are examples from 8 year olds.

So thank you for reminding me of all that! I am frustrated with the current shallow implementations of 'coding' I am witnessing in school systems. But, I'm still at it! As you all are! :-)

I'll now dig into the article more thoroughly.

~ Peter Skillen

---

Alan Kay

November 17, 2019 08:04

Hi Peter

All who knew Gavvy Salomon miss him very much. He had a unique way to present ideas, and was one of the most interesting interpreters of McLuhan.

Some in the ARPA/Parc research community noticed early on that when a human uses a tool, something also changes in the human's brain. McLuhan (and Innis, Havelock, et al) looked at this idea in the large, especially with regard to how we communicate ideas. They were sure that learning to read and write fluently was not just extending oral communication and oral societal thinking, but caused important qualitative changes in how thinking was actually done -- they strongly correlated all known civilizations with reading and writing.

This led to some deep visions and hopes for personal computing design: that if the approach was done as powerfully as reading and writing and learned as deeply, the results could be further giant leaps in how we can think about things. I still am quite convinced that this is the case.

However, there is now mountains of evidence that just learning to program does not produce these effects, and in fact seems to replace some richer modes of thinking with simpler more mechanical thought patterns. This -- like television -- has the overall effect of resetting "normal", in this case -- I think -- downwards.

The educational organizations in the world have so far done a very bad job of grappling with mathematics and most especially: science. And this after some hundreds of years of modern versions of these being in existence. We could at least say that "after WWI it would be a grave mistake not to have the average voting citizen be fluent in math and science".

So it's not surprising that they have quite flunked computing so far. This has greatly been complicated by the desire to replace real education with vocational training, and that the models of computing in business are quite poor. This produces a very dismal "normal"!

It has been one of the past main missions of public education to present "much better than normal" to children in the hopes and need that they will grow up with the discernment to raise normal above the important thresholds that we need today. It's hard to find this as a vision, mission, or goal in today's public education systems. Without something much better, we find that computing has turned into a kind of pop culture, with many of the same kinds of limited views as other "pop art forms".

---

Peter Skillen

November 20, 2019 10:25

Yes, Alan,

I very much agree with you. I am extremely discouraged by the 'pop culture' nature of computing and am very worried about the 'paint by number' nature of many of the 'coding' experiences to which children are exposed. Many educatorswell-intentionedexpect it to be a 'fun' experience for their students and for themselves but really aren't prepared to learn hard stuff. 'Hard fun' isn't part of their mindset.

I miss Seymour's voice these days. He is mentioned so much with this new crazebut, I fear his name is used superficially to further this paint by number approach. :-(

You say, "However, there is now mountains of evidence that just learning to program does not produce these effects, and in fact seems to replace some richer modes of thinking with simpler more mechanical thought patterns." Yeppers! In the 80s, there was enough research around to tell that story. Now, even more!

This desire for 'computational thinking for all' is misguided. Seymour did say it beautifully. I'll paraphrase not so beautifully. ;-) It is 'one' way of thinking and worth knowing. But, the biggest advantage is that the child can 'zoom out' and see it as ONE way while thinking about other ways of thinking. It becomes part of a repertoire of 'ways to think' that the child can draw on as required.

As Tom Lough used to say...FORWARD 100