What Happened to John Backus's Functional Programming Project?

John Backus, a pivotal figure in computing history best known for leading the development of Fortran, later became deeply dissatisfied with the imperative, 'word-at-a-time' style of programming inherent in the von Neumann architecture. This dissatisfaction culminated in his influential 1977 ACM Turing Award lecture, 'Can Programming Be Liberated from the von Neumann Style?', where he introduced his vision for a functional programming system and a new language called FP (Functional Programming).

Backus's FP language was designed to support a 'function-level' programming paradigm, emphasizing building programs from a fixed set of combining forms (functional forms) and avoiding named variables, a style often referred to as 'point-free' or 'tacit programming'. This approach aimed to create a more mathematical and provable style of programming, free from the complexities of mutable states and side effects that plagued conventional languages. FP was heavily influenced by Kenneth Iverson's APL, even adopting a non-standard character set.

Despite its theoretical elegance and the profound impact of Backus's paper, the FP language itself did not gain significant traction beyond academic circles. Critics noted that while it made complex operations easy, simple tasks could be surprisingly difficult to express, and its strict variable-free style was often less readable for many programmers. In the 1980s, Backus developed a successor language, FL (Function Level), as an internal IBM research project, but its development ceased upon project completion, and its source code was not publicly released.

However, the project's lasting consequence was its catalytic effect on the broader field of functional programming. Backus's lecture sparked renewed interest and research into functional paradigms, even though most subsequent functional languages, like Haskell and ML, evolved based on the lambda calculus rather than Backus's specific function-level style. These languages embraced core FP principles such as immutability, higher-order functions, and declarative programming, which Backus championed.

As of 2026, John Backus's original FP language remains largely a historical artifact, but his philosophical critique and the principles he advocated are more relevant than ever. Modern programming languages like Java, Python, and JavaScript are increasingly incorporating functional constructs, while purely functional languages like Haskell continue to innovate. Researchers are still exploring the potential of Backus's original formalism; for instance, a May 2026 paper discussed an open-source C++17 implementation of Backus FP, demonstrating its linear scalability on modern multicore systems and its potential for teaching parallel programming concepts. This shows a continued academic interest in the foundational ideas of Backus's project, adapting them to contemporary computing challenges.