The author points out that the history of APL is one of the removal of restrictions, citing as examples not having to declare the datatypes of language elements such as variables before use, not tying them to one datatype, and not having to declare array sizes. He goes on to propose the relaxation of another restriction in APL, the one that confines the use of parentheses to expressions that yield the argument of a function, and he gives examples of ways in which freer use of parentheses could contribute to the solutions to programming problems. He ends by proposing two further steps, allowing vectors of expressions as well as of functions, and removing the requirement that opening and closing parentheses in a single line must match. This, he argues, would allow “natural” (in APL terms) implementations of constructs familiar from other languages, if…else and case.

The paper will interest most regular and many occasional APL users, and those who do not routinely read APL conference proceedings or APL Quote Quad may wish to look it up. Syntax-mongering is a popular pastime in most language communities, and APLers are known for devising, by way of relaxation, programs that cram the maximum processing into one line; this paper may provoke further offerings of both kinds. The text is lively and readable but marred by a considerable number of misprints, the most prominent of which, the misspelling of “parentheses” in the running footer, is not the author’s fault. The misprints caused me to ponder a potential side-effect of removal of syntactic restrictions, namely the increased risk that a typo will produce, not an error, but the solution to a different problem from the one you thought you were programming.