A Model of the Babbage Firm

This paper develops a unified model of the cognitive division of labour in a knowledge economy. Building on recent frameworks for knowledge creation and decision making under uncertainty, it distinguishes between specialists, who engage in costly “on the spot” reasoning to generate tacit knowledge around a focal point, and generalists, who search for and interpolate existing knowledge but deliver answers subject to error. The model characterises how these two types of workers should be allocated across a continuum of questions, given the location of codified knowledge points and the distribution of problems. It shows that optimal task assignment depends on the cognitive process through which information is processed rather than on skill endowments or task complexity. When specialists operate around both knowledge points, their allocation is shaped by their absolute advantage over generalists, leading to non‐contiguous specialist domains interspersed with generalist regions. When specialists cluster around a single point, a natural boundary emerges between specialist and generalist domains that shifts but persists despite changes in question distribution. Extending the analysis to a two‐period setting, the paper identifies when specialists should sacrifice static efficiency to codify their tacit discoveries, creating bridges that allow generalists to operate more effectively in the future. These results provide a formal microfoundation for Babbage’s insights into the division of cognitive labour and offer predictions about how knowledge work responds to changes in the knowledge environment, the distribution of questions, and the patience of capital.