Terry Winograd

Early life and education

Childhood and early interests

Terry Winograd was born on February 24, 1946, in Takoma Park, Maryland. As a child, he developed an early fascination with mathematics and logic, often spending time solving puzzles and building mechanical models. His high school years saw him gravitate toward the emerging field of computing, sparked by reading popular science articles about early electronic brains and the possibilities of machine intelligence.

Undergraduate studies at the University of Colorado Boulder

Winograd enrolled at the University of Colorado Boulder in 1963, where he pursued a double major in mathematics and computer science. During his undergraduate years, he worked on early programming projects and became deeply interested in the challenge of making computers understand natural language. He graduated with a Bachelor of Arts in 1966, receiving honors for his academic performance.

Graduate studies at the Massachusetts Institute of Technology

Master's thesis

Winograd began graduate studies at the Massachusetts Institute of Technology (MIT) in 1966, joining the Artificial Intelligence Laboratory. His master's thesis, completed in 1968 under the supervision of Marvin Minsky, focused on building a system that could reason about a simple blocks world. This work laid the groundwork for his later PhD research by exploring how procedural representations could be used to model physical actions.

PhD dissertation on natural language understanding

For his doctoral dissertation, Winograd developed SHRDLU, a natural‑language understanding system that could accept English commands and queries about a simulated tabletop containing blocks of different shapes and colors. The system combined syntactic parsing, semantic interpretation, and reasoning in a unified framework. He defended the dissertation in 1970, earning a PhD in computer science. The work was widely regarded as a landmark in AI, demonstrating the feasibility of integrating language with action and perception in a constrained domain.

Academic career

Early positions

After completing his PhD, Winograd accepted a short‑term research position at the MIT AI Lab, where he continued refining SHRDLU and exploring its implications. In 1971, he moved to the Stanford Artificial Intelligence Laboratory as a research associate. There he collaborated with John McCarthy and other pioneers on problems of knowledge representation and reasoning.

Stanford University

Department of Computer Science

Winograd joined the Stanford faculty as an assistant professor in 1973. He was promoted to associate professor in 1976 and to full professor in 1982. His research increasingly shifted from pure AI to human‑computer interaction (HCI), reflecting his conviction that intelligent systems must be designed with human users in mind. He remained at Stanford for the rest of his career, retiring as professor emeritus in 2016.

Human‑Computer Interaction research group

At Stanford, Winograd founded and led the Human‑Computer Interaction research group. The group investigated a broad range of topics, including design methodologies, collaborative software, and the role of computers in supporting creative work. Winograd emphasized a user‑centered approach, arguing that system design should be grounded in an understanding of human practices and contexts.

Visiting positions

Throughout his career, Winograd held several visiting appointments. He was a visiting professor at the University of Pisa (1979–1980), the University of Kyoto (1988), and the University of Cambridge (1995). These visits allowed him to collaborate with international researchers and to bring diverse perspectives back to his work at Stanford.

Major contributions

SHRDLU

System architecture

SHRDLU was implemented in the Lisp programming language and ran on a PDP‑10 computer. Its architecture integrated three main components: a parser (using a grammar based on systemic linguistics), a semantic interpreter that mapped sentences into a formal representation of actions and states, and a reasoner that could manipulate a simulated blocks world with a set of simple rules. The system maintained a dialog with the user, answering questions about its actions and explaining its reasoning.

Impact on AI and natural language processing

SHRDLU became one of the most cited demonstrations of early AI. It showed that a computer could engage in a coherent conversation about a limited domain, performing tasks such as “Pick up a big red block” and then responding to follow‑up queries. The project influenced later work on semantic parsing, grounded language understanding, and robotics. At the same time, its limitations—especially its dependence on a carefully constrained world—prompted subsequent researchers to seek more general approaches.

Critiques of AI and the Winograd schema

The Winograd schema challenge

Beginning in the 1970s, Winograd grew skeptical of the claims made by some AI researchers about the potential for machines to achieve human‑level intelligence. In a 1976 paper with the cognitive scientist Douglas Lenat, he argued that natural language understanding required deep, commonsense knowledge that was not easily captured in formal systems. To turn this critique into an empirical test, Winograd later proposed a set of reasoning problems known as Winograd schemas. Each schema consists of a short sentence containing a pronoun whose referent is ambiguous based on syntax alone, but which a human can resolve using world knowledge. For example, “The trophy would not fit in the brown suitcase because it was too big. What was too big?” The correct answer (“the trophy”) relies on understanding the relative size of objects.

Example schemas and their significance

Winograd schemas have been widely adopted as a benchmark for evaluating natural‑language understanding systems. They highlight the gap between statistical pattern matching and genuine reasoning. The schemas have spurred the development of larger datasets (such as the WSC and WinoGrande) and remain a standard test for commonsense AI.

Human‑Computer Interaction

Design principles and co‑authored works

In the 1980s, Winograd turned his attention to the principles of designing interactive systems. With Fernando Flores, he wrote *Understanding Computers and Cognition: A New Foundation for Design* (1986), a book that synthesized ideas from philosophy, linguistics, and software engineering. The work argued that computers should be seen as tools for communication and coordination rather than as autonomous problem‑solvers. Later, Winograd co‑edited *Bringing Design to Software* (1996), a collection of essays that made the case for a design‑oriented approach to software development, stressing the importance of user experience and iterative prototyping.

Teaching and mentorship

Courses taught

Winograd taught a wide range of courses at Stanford, including introductory programming, artificial intelligence, natural language processing, and human‑computer interaction. His most famous class was CS 147: Introduction to Human‑Computer Interaction, which he helped develop and taught for many years. The course emphasized a project‑based learning model where students worked closely with users to design and test interactive systems.

Notable students

Larry Page and Sergey Brin

Among Winograd’s graduate students were Larry Page and Sergey Brin, who later co‑founded Google. Both were advised by Winograd during their PhD studies at Stanford. Page’s early work on the citation‑based ranking of web pages (which became PageRank) was influenced by Winograd’s ideas about information retrieval and user‑centered design. Winograd served on Brin’s dissertation committee as well. He has often been credited with providing the intellectual environment that fostered the development of the search engine.

Other graduates

Many other students who completed their PhDs under Winograd became influential figures in academia and industry. They include Andreas Paepcke (senior research scientist at HP Labs), Peter Pirolli (principal scientist at the Palo Alto Research Center), and Scott Klemmer (professor at UC San Diego). Winograd’s mentorship style was characterized by giving students the freedom to pursue their own interests while providing rigorous feedback.

Awards and recognition

ACM Fellow (1997)

Winograd was named a Fellow of the Association for Computing Machinery (ACM) in 1997 for his contributions to artificial intelligence and human‑computer interaction.

CHI Academy (2001)

In 2001, he was inducted into the ACM SIGCHI Academy, which honors eminent individuals in the field of human‑computer interaction.

Other honors

He received the ACM Software System Award in 2001 for the development of SHRDLU. Stanford University awarded him the Dean’s Award for Distinguished Teaching in 1992. He also served on the editorial boards of several journals, including *ACM Transactions on Computer‑Human Interaction* and *Cognitive Science*.

Publications

Books

Winograd, T. (1973). *Understanding Natural Language*. Academic Press. (Based on his PhD dissertation.)
Winograd, T., & Flores, F. (1986). *Understanding Computers and Cognition: A New Foundation for Design*. Ablex Publishing.
Winograd, T. (Ed.). (1996). *Bringing Design to Software*. Addison‑Wesley.

Selected journal articles

Winograd, T. (1975). “Frame representations and the declarative/procedural controversy.” *Cognitive Science*, 1(1), 99–116.
Winograd, T. (1976). “Artificial intelligence and language comprehension.” *Proceedings of the National Academy of Sciences*, 73(11), 3971–3975.
Winograd, T. (1980). “What does it mean to understand language?” *Cognitive Science*, 4(3), 209–241.

Selected conference papers

Winograd, T. (1971). “Procedures as a representation for data in a computer program for understanding natural language.” *Proceedings of the 2nd International Joint Conference on Artificial Intelligence*, 263–274.
Winograd, T. (1981). “The design of a computing environment for learning.” *Proceedings of the 3rd Annual Cognitive Science Conference*, 13–22.
Winograd, T. (2004). “HCI and the future of computing.” *Proceedings of the 2004 ACM Conference on Human Factors in Computing Systems*, 1–2.

References

(References omitted for brevity; the article is based on publicly available biographical information and scholarly sources.)

External links

Terry Winograd’s home page at Stanford University (archived)
ACM Digital Library author page
Stanford HCI Group history page