Conceptual Dependency theory is Roger Schank’s proposal that the meaning of a sentence can be represented in a canonical form built from a small fixed set of primitive acts, independent of the particular words used. It was developed through the late 1960s and 1970s and is laid out, together with the scripts and plans that build on it, in Schank and Abelson’s 1977 book “Scripts, Plans, Goals and Understanding,” whose full text is preserved by the Internet Archive and cited here.
The central claim is that beneath the surface variety of language lies a much smaller vocabulary of conceptual primitives. Acts like physically transferring an object, transferring possession, or transferring information between minds are written with primitives such as PTRANS, ATRANS, and MTRANS. So “John read a book” is represented as John MTRANS-ing information from the book into his memory. The point of this reduction is paraphrase independence: two sentences that mean the same thing, even in different languages, should map to the same conceptual structure, so a program can recognize that they say the same thing and can draw the same inferences from either. The representation also makes implicit consequences explicit, supporting the inference needed to understand connected text rather than isolated sentences.
Conceptual Dependency was the foundation for a line of Yale story-understanding and machine-translation programs, and it pushed the field toward meaning-based, rather than purely syntactic, approaches to language. Its ambition, a universal, language-neutral representation of meaning, anticipated later interlingua and semantic-representation efforts.
Why a business reader should care: the dream of capturing what a text means in a structured, language-independent form, so software can compare, query, and reason over content rather than just match words, runs straight from Conceptual Dependency to today’s knowledge graphs and semantic search.