# John McCarthy

> 1927–2011 · Computer Scientist, Creator of Lisp
>
> **Recorded contribution:** Coined "artificial intelligence"; LISP; garbage collection; time-sharing; situation calculus

## How to use this dossier

Read for a causal chain, not a hero story: inherited problem → contribution → mechanism → downstream capability → limit. Then close the page and complete the reconstruction exercise from memory.

## 1. Historical orientation

John McCarthy (1927–2011) named 'artificial intelligence' in the 1955 Dartmouth proposal and built several of the field's durable abstractions. At MIT he designed Lisp around symbolic expressions, recursion, and functions; he also promoted time-sharing and later developed situation calculus and nonmonotonic reasoning for common-sense knowledge. Lisp's original evaluator made program and data share a uniform list representation, enabling interactive metaprogramming and rapid AI experimentation. McCarthy did not single-handedly invent AI, garbage collection, or time-sharing: each emerged from communities and precedents. His distinctive power was to formulate research problems and computational languages in unusually general terms.

## 2. The problem inherited

AI programs needed to construct, traverse, and transform symbolic expressions whose sizes and shapes change dynamically, but machine-oriented languages made this cumbersome.

## 3. The central contribution

McCarthy designed Lisp and framed AI as the study of machines performing intelligence-requiring tasks, supplying a programmable medium for recursive symbolic computation and research.

## 4. Reconstruct the mechanism

1. Represent both code and symbolic data as atoms and linked lists expressed with the same S-expression notation.
2. Evaluate an expression by inspecting its operator, evaluating arguments as required, and applying a primitive or user-defined function.
3. Use recursion and conditional expressions to traverse arbitrary symbolic structures without fixed array bounds.
4. Allocate list cells dynamically and reclaim unreachable cells, allowing long interactive sessions to create new structures.

## 5. What changed downstream

- Lisp became a foundational AI language and influenced functional programming, dynamic languages, and garbage-collected runtimes.
- The Dartmouth project gave a diverse research agenda a memorable institutional label.
- Situation calculus and related work shaped formal knowledge representation and reasoning about action.

## 6. Attribution, limits, and uncertainty

- The Dartmouth proposal had McCarthy, Minsky, Rochester, and Shannon as authors and built on logic, cybernetics, neuroscience, and earlier computation.
- Lisp implementation depended on Steve Russell and other MIT colleagues; garbage collection also has a broader development history.
- Formal common-sense representations face ambiguity, incomplete knowledge, grounding, and combinatorial-scale problems.

## 7. Reconstruction lab

Implement a minimal Lisp evaluator for quote, atom, eq, cons, car, cdr, cond, lambda, and function application. Use it to write recursive append, then mark reachable list cells from an environment and sweep the rest as a simple garbage collector.

## 8. Evidence trail

- [John McCarthy, 1927–2011](https://news.stanford.edu/stories/2011/10/john-mccarthy-obit-102511) — Stanford University
- [Recursive Functions of Symbolic Expressions and Their Computation by Machine](https://doi.org/10.1145/367177.367199) — Communications of the ACM

---

*Research checked 2026-08-09. Dates, roles, and claims about living people are historical snapshots. Linked sources remain the authority; this dossier is original instructional synthesis.*