4.24: Second Vision: Apple II with Logo

This vision is, perhaps, best introduced, somewhat whimsically, by telling a story of three Wise Men bringing gifts to a child - or to anyone who has ever been a child.

The first Wise Man is Jean Piaget. His gift was a theory. His theory does for ontogenetic development (from child to adult) what Darwin's theory does for phylogenetic development (from animal to human). He has provided us with a three-dimensional description of human development in all its length, breadth, and depth.12

The second Wise Man is Seymour Papert. His gift was a language. A student of Jean Piaget, he and his colleagues at the Massachusetts Institute of Technology (MIT) developed a computer language called Logo, which would enable computers to facilitate human development, as described by Piaget.13 Logo was adapted from LISP, a computer language used at MIT to study artificial intelligence, as a language for learning how to learn.

This language was first "taught" to a mechanical "Turtle" attached to a keypad. A child (of whatever age) would type FORWARD 40 and the Turtle would take 40 turtle steps forward, leaving a trace of its trip with a pen on a paper on the floor. The child may then type RIGHT 90 and the Turtle would turn 90 degrees to the right. Using four basic words in the Logo language (FORWARD, BACK, RIGHT, LEFT), the Turtle can be instructed to draw any shape on the paper.

Let us imagine that a child has commanded the Turtle to draw a triangle, then a square, then a pentagon, and so on. The child may already have discovered the Total Turtle Trip Theorem - that is, regardless of the shape of the figure, the turtle must turn a total of 360 degrees to get back to the position in which it started. This is only one of many insights into geometry the child could get using only those four simple words.

Let us say that the child now wishes to teach the Turtle how to draw a circle. The Logo teacher may encourage the child to "think Turtle" This could help the child realize that the Turtle will have to go forward a little, turn a little, many times. Few children would have the patience to key in FORWARD 1, RIGHT 1 360 times. No child would be willing to repeat this chore every time he/she wanted the Turtle to draw a circle. This brings us to the third Wise Man.

The third wise man is Guy Montpetit. His gift is the distribution to the general public of a disk containing the Logo language. A student of both Piaget and Papert, he returned home to Montreal to found Logo Computer Systems Inc. (LCSI) to manufacture and distribute software using the Logo language. The mechanical Turtle on the floor may thus be replaced by a triangle on a screen and the keypad by a keyboard. A disk, supplied by LCSI, could be considered as the "mind" of the Turtle. It mediates between you at the keyboard and the Turtle-triangle on the screen.

It solves the two problems mentioned above. The first problem of repeating FORWARD 1 RIGHT 1 360 tedious times is solved by simply typing REPEAT 360 (FORWARD 1 RIGHT 1). REPEAT is another Logo word which the Turtle "understands" because it is programmed into the disk.

The second problem of repeating this command every time you want a circle is solved by teaching the Turtle a new word. This is done as follows:

TO CIRCLE tells the turtle that you are about to teach it a new word, which will be called CIRCLE. REPEAT 360 (FORWARD 1 RIGHT 1) teaches it the "meaning" of the word CIRCLE. END tells the Turtle that you have finished defining the new word. (Thereby following an important teaching principle: you tell them what you are going to tell them, you tell them, you tell them what you just told them.) Now, when you want circle, you simply type CIRCLE.

One of the best ways to understand the relatively unfamiliar concept of a computer language (Logo, BASIC, etc.) is by analogy with the more familiar concept of a natural language (English, French, etc.).

A natural language, the linguists tell us, consists of a hierarchy of units and sets of rules for combining acceptable units at one level to create meaningful units at the next level. What's that? Linguists don't talk too clearly. Figure 4-2 may help make this more concrete. A natural language consists of phonemes (roughly corresponding to letters) which can be combined, according to the rules of vocabulary, to yield morphemes (roughly corresponding to words) which can be combined, according to the rules of grammar, to yield sentences which can be combined, according to the rules of logic, to yield discourses.

The structure of a computer language is essentially the same. Thus, learning a language, whether natural or artificial, consists largely of learning those acceptable units and the appropriate rules of vocabulary, grammar, and logic to combine them into units at the next level.

Though they have essentially the same structure, there are some interesting differences between Logo and English.

  • Logo is an artificial rather than a natural language.
Logo was developed mechanically from the outside in, whereas English grew organically from the inside out. It was consciously developed for a specific purpose - to talk to computers, or, more accurately, to talk to computer programs or, more whimsically, to talk to Turtles or to Turtles which had been transformed into triangles.

  • Logo is typed rather than talked.
It is not possible to talk to computers, computer program, or Turtle-triangles (at least, not yet). You can't yet write to them either (you can, if you like, but you won't get an answer). You can, however, type to them and they can type back. The set of characters (corresponding to the phonemes in a natural language) are all on the keyboard. The Logo language is designed so that you can talk-type on the keyboard to the Turtle, represented by a triangle on the screen, and the Turtle can talk-type back to you on the screen.

  • You can invent new Logo words.
The Turtle has a basic vocabulary of Logo words, which it can understand. You could think of them as the instincts of the Turtle. However, you can teach the Turtle new words (as we saw above in teaching the Turtle the meanings of SQUARE and CIRCLE). You can, of course, invent new words in English. However, they will be of no value in communication, except with those people to whom you teach their meaning.

  • Your sentences in Logo tend to be commands.
Sentences in English can be roughly classified as statements (followed by a period), questions (followed by a question mark) and commands (followed by an exclamation point). When you talk Logo to the Turtle, you tend to use commands and the Turtle tends to use questions. It is like a conversation between a Sergeant-Major and a Private.

  • Logo is more precise and concise than English.
When talk-typing to the Turtle, you must talk very precisely. The Turtle is not very smart. It will not understand you unless you type exactly what you want. You can also talk very concisely. Most Logo words have a shorthand version - e.g. FORWARD can be written FD and RIGHT can be written RT. You can, of course, do the same in English (e.g. TELEVISION can be written TV and MOdulator-DEModulator can be written MODEM). However, since it takes longer to type than to talk, this practice is more common in Logo than in English.

Logo differs, in a number of ways, from various traditional computer languages (FORTRAN, COBOL, BASIC, etc.). Those languages were created at a time when computer memory was very expensive. They are therefore designed to conform to the requirements of the machine rather than of the person. BASIC is a brave attempt to consider the needs of the person my simplifying FORTRAN so that it is easy to learn. It is not, however, easy to use. Nor is it powerful enough for complex programs. Thus, one must learn BASIC to do simple programming and, then, learn FORTRAN to do complex programming.

Logo, on the other hand, was specifically designed to help children learn to learn. It is simple yet powerful. Thus, it has a very low threshold and a very high ceiling. Turtle Geometry, an undergraduate textbook, starts with a first few fumbling turtle steps and ends with Einstein's General Theory of Relativity.14 At first, Logo could be used only in laboratories with very large computers. However, now that home computers are becoming smarter and smarter and cheaper and cheaper, Logo can be made available to the general public.

A language like Logo, which is based on child development, is easier to learn and to use than languages like FORTRAN, which are based on engineering. It is designed for people rather than for machines.

There is, however, another reason why it is better to "talk" Logo rather than FORTRAN. Benjamin Lee Whorf has argued that the language we speak determines the way we think.15 Though this theory is very controversial, there is considerable evidence that language determines thought, in some aspects and to some extent. Perhaps, the computer language we use also determines the way we think.16 A language designed for people would, therefore, have a more congenial effect on thought than a language designed for machines.

Despite those advantages of high-level languages like Logo, there is a danger that we will succumb to the "qwerty phenomenon".17 The traditional keyboard (called "qwerty" after the first six letters on the top left side of the keyboard - not after the inventor, Mr. Qwerty, as I first thought) was designed by mixing up the letters randomly to slow down the typist because the keys in the early typewriters would jam if the operator went too fast. The keys didn't stick but the keyboard did. We are stuck with it. There are much more logical keyboards available, but the qwerty keyboard can not be replaced, because so many people have been trained on it. Traditional computer languages may be to the computer what the qwerty keyboard is to the typewriter.

12   Flavell, J. H. Developmental Psychology of Jean Piaget. New York: Van Nostrand Reinhold, 1963.

13   Seymour Papert, Mind-storms: Children, Computers, and Powerful Ideas. New York: Basic Books, 1980.

14   Abelson, Harold & Andrea diSessa, Turtle Geometry: Computation as a Medium for Exploring Mathematics. Cambridge, Mass.: MIT Press, 1982.

15   J. B. Carroll [Editor]. Language, Thought and Reality: Selected Writings of Benjamin Lee Whorf. Boston: MIT Press, 1956.

16   My thanks to my friend, Gordon Sheppard, for this idea.

17   Seymour Papert, Mind-storms: Children, Computers, and Powerful Ideas. New York: Basic Books, 1980.