Some of it depends on perspective. I was introduced to Logo in 1983 on a Commodore 64, and got a lot of mileage out of turtle graphics. (To the point of experim

Some of it depends on perspective. I was introduced to Logo in 1983 on a Commodore 64, and got a lot of mileage out of turtle graphics. (To the point of experimenting with recursion and circles, etc.) The other thing to realize about Logo is that it was developed as a version of Lisp suitable for teaching programming concepts to young children. Turtle graphics were added because it was very anthropomorphic - children could more easily do things like give turtle graphics style instructions to other children and visualize themselves in the place of the turtle to debug their problems. It was intended to be close to a 'zero-entry swimming pool' style approach to programming. But turtle graphics were intended to be only the beginning. The language also has a full set of list and structured data primitives. Children that started learning how to draw squares and geometry on the screen could take that basic skillset and use it to do progressively more advanced computation. Commodore 64 Logo included a version of the Animal guessing game that was built on List manipulation and didn't use turtle graphics at all. That was directionally where learners were expected to take the language, given enough time and appropriate education. Underpinning all of this was a belief in constructivist education that had kids learning through doing, building, and experimenting. The idea was that Logo was a way to facilitate this sort of thing in an accessible way. How often do children (or anybody) get to define the rules of a 'world' and conduct experiments to see what those rules produce. There's a lot of idealism in all of this, but I consider myself very fortunate to have been exposed to it when I was.