Logo: Classroom Prototypes and Turtle Programming

Logo's design began in the 1960s at Bolt Beranek and Newman, usually abbreviated BBN, in Cambridge, Massachusetts. The team wanted children to explore mathematics by writing, running and revising programs, rather than merely sitting at terminals answering a machine's questions. Programming was both a means of expression and an activity through which ideas could be tested.

The 1969 project report credited the original language design to Seymour Papert, Daniel Bobrow and Wallace Feurzeig, noting Cynthia Solomon's participation in developing and using the language and Richard Grant and others' continuing implementation work. In their 2020 history, Solomon and colleagues recalled that design discussions had begun in 1966. They drew on Lisp's symbolic-processing capabilities, especially the ability for children to manipulate words and sentences rather than start with algebraic notation.

References: [2] [1]

According to Solomon's participant recollection, Bobrow first implemented the new language in Lisp on BBN's SDS-940 time-sharing computer, with colleagues continuing the work. Feurzeig named it Logo, relating the name to the Greek word for “word”. By summer 1967, a functioning system was being tried by children at Hanscom School in Lincoln, Massachusetts. Papert taught, and the team discussed and revised the system before and after classes.

The year 1967 thus marks an early stage when the prototype entered classrooms. The version prepared for the 1968–1969 school year moved to a DEC PDP-1, with school terminals connected to BBN's time-sharing computer. Design, experimental teaching, language revision and platform changes intertwined; they cannot be compressed into one mass-market product launch. Programs needed to run, and the language also had to reveal its problems through children's actual use.

References: [1]

The 1969 report used DOUBLE to explain procedure design: adding a number to itself yielded twice its value. The initial procedure merely printed the answer on the teletype. Entering 4 displayed 8, but that 8 could not be passed onward to an enclosing procedure. Only changing PRINT to OUTPUT made the result a value other procedures could use.

PRINT DOUBLE OF (DOUBLE OF 4) would then print 16: the inner procedure first changed 4 to 8, and the outer changed 8 to 16. The example distinguished a display for a person from a value for a program and showed how a defined procedure could become a building block for the next step. A learner who only checked whether an answer appeared on paper would miss why composition failed. Running and revising made abstract ideas of input, output and function composition observable.

References: [2]

In September 1968, an experimental Logo class began at Muzzey Junior High School in Lexington, Massachusetts, with twelve seventh-grade students. Marjorie Bloom initially taught it, and Solomon and Papert continued jointly from January to June the following year. The formal report submitted to the U.S. National Science Foundation on November 30, 1969 included classroom materials and descriptions of the language and system, preserving technical texts for others to inspect.

The report discussed standardized testing, changes in student grouping and comments from visiting mathematics educators, while acknowledging the small sample. The existence of a course, children's completion of programs and changes in a test result were different levels of evidence. Interpreting an exploratory classroom as proof that any school adopting Logo would improve achievement would exceed what this study could establish.

References: [2]

In Twenty Things to Do with a Computer, issued by the MIT Artificial Intelligence Laboratory in 1971, Papert and Solomon extended applications to mechanical turtles, graphics, music and other controllable devices. Their turtle connected by cable to control equipment, could move and turn, and received sensor signals. Lowering a central pen made its movement leave a trace. This was a further development after early text processing and classroom experiments.

In the report, FORWARD 50 moved the turtle in its current direction, while LEFT 90 changed its heading without changing its position. Understanding those two states explained why the same distance command could draw in different directions. The report also distinguished floor turtles from screen turtles: the former encountered a real environment, while the latter could quickly draw and erase graphics. Both made geometric relationships executable actions, but they were not one unified feature list of the 1967 prototype.

References: [3]

With personal computers, Logo no longer depended exclusively on research institutions' time-sharing equipment. The 2020 history recorded Apple II implementations in 1981, including Terrapin Logo and Apple Logo from Logo Computer Systems. Later versions continued procedure composition and interactive experimentation while diverging according to machine resources, development teams and intended uses.

Independent research by Angelos Agalianos, Richard Noss and Geoff Whitty used interviews and classroom observations to examine Logo in British and American schools. Open-ended projects could require teachers and students to debug together, yet faced fixed lesson periods, subject-based curricula and examination pressures. Some schools turned Logo into a separate computing lesson or reduced it to copying drawings. The researchers stressed that these changes could not simply be blamed on individual teachers.

As Logo moved from prototype to wider adoption, what spread was not only a set of commands but also expectations of classroom organization. Its historical questions are therefore concrete: who decides what students make, how errors are discovered, and whether finishing a picture leads to further explanation of the program. Answering requires understanding implementation, teaching materials and school contexts together.

References: [1] [4]