In a retrospective look at APL, a programming language that has often been overshadowed by more modern counterparts, Professor Alan J. Perlis of Yale University emphasizes its unique qualities that make it an ideal first language for teaching computer science. Published in 1977, Perlis's article 'In Praise of APL' argues that APL's terseness, flexibility, and composability offer unparalleled benefits for students just beginning their journey into programming.

Perlis contends that APL allows students to express complex ideas in a concise manner, fostering a deeper understanding of algorithmic thinking. Unlike BASIC or FORTRAN, which were easier to learn initially, APL's rich set of primitive functions and its ability to handle arrays make it a more powerful tool for developing a wide range of programs. This language encourages students to think creatively and develop an arsenal of idioms that become second nature, much like natural language prose.

The article also highlights APL's role in teaching computer organization, where its ability to model computer components at various levels of detail makes it a practical choice. With APL, students can explore topics such as word processing, graphics, and file processing without being overwhelmed by the complexity of pre-packaged systems. This hands-on experience empowers students to create their own systems, fostering a deeper understanding of the man-machine symbiosis and encouraging them to adapt tools to their specific needs.

As the field of computer science continues to evolve, revisiting languages like APL offers insights into the fundamental principles of programming and algorithmic thinking. While newer languages have emerged, the principles that made APL a strong choice for teaching remain relevant. Understanding these principles can help educators and students appreciate the evolution of programming languages and the enduring importance of clear, concise, and flexible coding practices.