Misconceptions about Logo

Misconceptions about Logo by Seymour Papert.

People often congratulate me for making such a good language for children. But they are wrong. Logo isn’t a “good language for children”–in fact, a language that was “good for children” in this limited sense would not be good for children. Children deserve a language that is good, period. Logo is only good for children insofar as it is good for everyone.

Something similar can be said in response to other remarks one hears about Logo. They say it is a good language for graphics. But this is also looking at things the wrong way around. A language that was “good for graphics” and nothing else would not be a good language for graphics. To do graphics well, you need a powerful, general purpose language.

Why is something that is intended to be “good for children” not really good for children? Some obvious examples are all the simplifications and perversions of Logo, the so-called “Instant Logo” that springs to mind. Some of the commercial products are also simplified versions of the sorts of things children usually do in their first few days, perhaps even weeks, of Logo.

The reasoning behind these Instant versions is that for a young child who does not yet know the alphabet and the keyboard, typing the command FD 10 is too time-consuming and frustrating. So these six keystrokes (including the space and carriage return) are reduced to one: F. But this robs the child of the opportunity to think about the experiment with numbers that is implicit in Logo. This restriction becomes even more significant in exploring angles. How can one discover what a 45-degree angle looks like when he can turn the turtle only in multiples of 10 or 30 degrees with each keystroke? Or that LEFT 90 and RIGHT 270 produce the same result?

Far more important, Instant versions of Logo don’t provide the experience of gaining an increasingly rich understanding of the same intellectual entity. An exciting educational aspect of Logo is that a five-year-old can do something interesting with it, and a graduate student can do something else interesting with it. The fact that each is using the same language, exploring the same system, means that both have the opportunity to gain a richer and deeper understanding over a long period of time–albeit at different levels–of materials that are essentially continuous.

This brings us to another misconception. Logo is said to be an “easy” language–but it is not. It is designed to have easy routes into the language. Using commands like FORWARD 20 and RIGHT 90, one can do quite interesting work at this level of comprehension–where Logo statements sound like English and can be understood in this spirit. But there comes a time when more subtle kinds of understanding are necessary. No child (nor any adult who is not sophisticated about formal systems) get there without long experience or good instruction, because these more subtle ideas really are difficult. Logo would have no point if it were nothing but an easy language. It has a point because it starts off easy and then becomes difficult gradually enough that no one need ever drop out.

In this respect, Logo is much like a natural language such as English. Baby talk allows the child an easy way to grab onto and use pieces of the language. But nowhere in the world is a language restricted to baby talk. The child continues to be immersed in the full richness of language as used by philosophers and poets. It is the quality–the rich complexity that opens up as one explores Logo ever more deeply–that makes Logo “good for children” and for adults as well.

Papert, S. “Misconceptions about Logo” in Creative Computing. Vol. 10, No. 11 / November 1984 / PAGE 229. Accessed via this web page.