Cognition in Practice

In the AI Story

The book arrived at a moment when cognitive science was dominated by the information-processing paradigm — the view that mind is computation, knowledge is representation, and cognition can be studied in laboratory isolation from the contexts in which it normally occurs. Lave's challenge was not that this paradigm was internally inconsistent but that it systematically misrepresented how cognition actually works in the world. The book marshals extensive ethnographic evidence to demonstrate that cognition in the wild looks radically different from cognition in the lab — and that the laboratory version is not a purified version of the wild phenomenon but a different phenomenon altogether.

The theoretical framework Lave develops in the book draws on Marxist and activity-theoretic traditions, particularly the work of Soviet psychologists like Lev Vygotsky and Alexei Leontiev, as well as on the sociology of Pierre Bourdieu and the phenomenology of Maurice Merleau-Ponty. From these sources Lave synthesizes a position that cognition is dialectical — constituted through the ongoing interaction between the person and the setting, neither reducible to the individual mind nor determined by external structure.

The book's most influential contribution is probably its methodological demonstration that ethnography can produce evidence about cognition that experimental psychology cannot. The Adult Math Project data, presented in Chapter 4, shows with quantitative precision what ethnographic observation had suggested qualitatively: that performance in everyday settings bears little resemblance to performance on formal tests of the same mathematical operations. The data makes the theoretical claims difficult to dismiss.

For readers encountering the book in the AI era, it reads as an extended meditation on exactly the question the current moment forces: what is the relationship between decontextualized information and situated understanding? Lave's answer — that the two are different phenomena produced by different processes, and that the situated kind cannot be produced by exposure to decontextualized information alone — has acquired new stakes as large language models industrialize decontextualization on a scale Lave could not have anticipated.

Origin

The book was published by Cambridge University Press in 1988 and developed from Lave's ethnographic research in the late 1970s and early 1980s, conducted primarily in Orange County, California. It built on her earlier fieldwork among Vai and Gola tailors in Liberia in the 1970s, and laid the theoretical groundwork for Situated Learning (1991, with Etienne Wenger).

Key Ideas

Cognition is dialectical. It emerges from the ongoing interaction between persons and settings, not from mental operations alone or from external structures alone.

Everyday arithmetic is not impoverished school arithmetic. It is a different cognitive achievement, using different resources, operating under different constraints, and often producing more reliable results than its formal counterpart.

Settings are not incidental. The physical arrangement of products, the social context of the activity, the practitioner's embodied history — these are constitutive of the cognitive performance, not external to it.

The laboratory is a specific cultural setting. Its findings do not generalize to other cultural settings because they are the findings of one cultural setting — the setting in which people perform cognition for psychologists.

Debates & Critiques

The book was initially received with skepticism by mainstream cognitive psychology, which questioned its generalizability and its theoretical ambitions. Over the subsequent three decades, its central claims have been substantially vindicated by research across many domains — but the practical implications for educational design remain contested. Reformers inspired by Lave's work have advocated project-based learning, apprenticeship models, and community-engaged pedagogy. Critics argue that these approaches, while sometimes effective, cannot scale to the mass educational systems that modern economies require.