Introduction. March’s 1991 paper frames organisational learning as an allocation between exploration (search, variation, experimentation, whose returns are distant and uncertain) and exploitation (refinement, efficiency, whose returns are near and reliable). Two models carry the argument. In the first, individuals learn from an organisational code and the code learns from the individuals who outperform it; fast learners converge on the code quickly and the code stops improving, whereas slow learners and turnover preserve the diversity the code needs. In the second, competition for primacy rewards variance rather than mean performance. The central result is that adaptive processes refine exploitation faster than exploration, so they are effective in the short run and self-destructive in the long run. The competency trap (Levitt and March) is the mechanism: experience with an inferior procedure raises its returns enough that a superior one is never tried. Later work located the balance in structure: Lazer and Friedman showed that efficient communication networks converge faster and reach worse solutions; Fang, Lee and Schilling showed that semi-isolated subgroups with occasional cross-links outperform both fully connected and fully isolated designs; the ambidexterity literature asks whether one organisation can do both.
Important authors. James G. March (1928–2018), Stanford, earlier at Carnegie Tech with Simon and Cyert, founded the behavioural theory of the firm and the study of organisational learning under ambiguity. Barbara Levitt co-authored the competency-trap paper; Daniel Levinthal (Wharton) co-authored “The myopia of learning” and carried the theme into landscape models; David Lazer (Northeastern) and Melissa Schilling (NYU Stern) produced the network results; Charles O’Reilly (Stanford) and Michael Tushman (Harvard) lead the ambidexterity line.
Importance for cybernetics and the VSM. The exploration–exploitation trade-off is a rival to the Three–Four homeostat: where the VSM treats the balance between present operations and future adaptation as something a designed homeostat can hold, March shows that ordinary adaptive learning drifts toward exploitation on its own, and that the drift is accelerated by fast convergence — which is what effective coordination produces. The VSM’s tacit “more coordination is better, up to the autonomy constraint” reading of System Two therefore has a documented adverse consequence: premature convergence and competency traps, neither of which has a VSM name.
Importance for the article. §3.3’s pathology list opens with “competency traps (March, 1991†),” and the candidate receipt in §10.5 — “More coordination (System Two) is better, up to the autonomy constraint” against “optimum connectivity is interior and horizon-dependent” — pairs March’s mechanism with Derex and Boyd’s experiment as the discriminator. The import is Applying new lenses (C5 §IV and §VI: “premature convergence on a local optimum, which the VSM’s coordination machinery would accelerate”). March is Tier C, marked †. A reviewer will press on whether the VSM in fact prescribes maximal coordination (Beer’s autonomy principle says otherwise) and whether the two readings predict different data. The answer is that the receipt targets the tacit teaching, not Beer’s text, and that the discriminator — cumulative-innovation experiments with VSM-specified coordination regimes — predicts an interior optimum for System Two connectivity that shifts with time horizon. Fang, Lee and Schilling is the closest organisational evidence and belongs beside Derex and Boyd.
Sources in the reading list.
- the mutual-learning model (section on learning rates and turnover) for the mechanism the §10.5 receipt relies on.
Other important sources and authors.
- Cyert, R. M., & March, J. G. (1963). A Behavioral Theory of the Firm. Prentice-Hall — the source of problemistic search and organisational slack.
- Levitt, B., & March, J. G. (1988). Organizational learning. Annual Review of Sociology, 14, 319–340 — the competency trap defined.
- Levinthal, D. A., & March, J. G. (1993). The myopia of learning. Strategic Management Journal, 14(S2), 95–112 — temporal, spatial and failure myopia; the failure myopia is survivorship again.
- Lazer, D., & Friedman, A. (2007). The network structure of exploration and exploitation. Administrative Science Quarterly, 52(4), 667–694 — efficient networks converge on worse solutions; the simulation counterpart of Derex and Boyd.
- Fang, C., Lee, J., & Schilling, M. A. (2010). Balancing exploration and exploitation through structural design: The isolation of subgroups and organizational learning. Organization Science, 21(3), 625–642 — semi-isolated subgroups as the design solution; the organisational version of interior optimum connectivity.
- O’Reilly, C. A., & Tushman, M. L. (2013). Organizational ambidexterity: Past, present, and future. Academy of Management Perspectives, 27(4), 324–338 — the review of whether both can be held in one structure.