Example · part of Thinking inside frameworks

Show a Pirahã speaker a set of objects and ask for a matching set. The Pirahã language has no stable word even for “one.” Yet when the objects stayed in view, Pirahã speakers matched large exact quantities perfectly. When the task required memory, they were inaccurate.

Frank, Everett, Fedorenko and Gibson reported this in 2008 and drew a careful conclusion. Number words do not change how people represent number. They are a “cognitive technology” for keeping track of exact amounts across time, space and changes of form.

So number words are a tool that extends a capacity. The book treats vocabulary in general this way: shared words stretch the receiver budget, the limited capacity each person has for taking things in, and they bear on whether that budget is fixed or trainable. Written numerals do similar work for arithmetic, the point of notation as a tool of thought.

This is one study of one community, and it cuts against the strong form of linguistic relativity. Lacking number words did not stop speakers from handling exact quantity. It made one job, holding a quantity in memory, harder. The Russian blues study shows the same limit: influence, not imprisonment.

How solid is this? One published study (Cognition, 2008); it supports the weak view of language as a tool and counts against the strong view that a language fixes what its speakers can think.

Read the full argument

How it connects


← Russian blues · Thinking inside frameworks (10 of 21) · Influence, not imprisonment →

Explore all the ideas · Read the book in order · Open the interactive map