What Cognitive Abilities the Science Tells us to Seek

The interviews break Meadows' practice into tasks and check each task against cognitive research. They map tasks, so they say nothing about where in the brain system-thinking happens, or whether it lives in a single network.

Meadows, in the form we can test

The phenotype comes from Thinking in Systems (2008), from Dancing with Systems, and from Leverage Points. A competent thinker watches a trajectory before acting, keeps a stock distinct from its flows, treats a loop as circular cause with a delay, recognizes the same skeleton under different nouns, ranks a goal or an information link above a parameter, and retires the model when a fact kills it.

In her writing, humility is something you do. We score it only when it shows up as a falsifier or a revised structure, never as a personality trait.

Four candidate capacities

The project asked whether exceptional systems thinkers are people with unusually large working memory, rich semantic memory, strong spatial or 3D imagery, and superior pattern recognition. The literature splits that package.

Dynamic tasks

Sterman (1989) showed managers in a delayed inventory system under-using feedback from their own decisions. Gonzalez, Fakhari, and Busemeyer (2017) review learning in dynamic decisions, where the next state depends on the last action. Fleming (2026) reviews metacognition as the use of uncertainty. We borrow the behavioral lessons and claim no neural signature for the interview.

Lalani, Gray, and Mitra-Ganguli (2023) is the nearest paper that says "systems thinking" and "cognitive neuroscience" in one title. It is a perspective piece on climate-smart agriculture that suggests where one might look; it reports no localization result.

Robustness, the property Meadows said is hard to see

Meadows already made room for it. In chapter 3 of Thinking in Systems she writes that systems "need to be managed not only for productivity or stability, they also need to be managed for resilience", and she notes that people often sacrifice it for stability or productivity, because resilience is hard to see without a view of the whole system. The earlier sections ask whether someone can read how a system behaves. This one asks whether they can judge what it will survive.

Olivier Hamant, a plant biologist, argues from living systems that their stability comes from the inefficiencies an engineer would remove: redundancy, heterogeneity, slack, slow processes, cooperation, and constant small adjustments. His 2024 paper in Quantitative Plant Biology presents this as "a call for systems thinking". Kitano (2004) reached a similar conclusion in molecular biology and adds the warning that matters for us: there are "trade-offs between robustness, fragility, performance and resource demands", so a design tuned against the usual shocks can be fragile to an unusual one. Ulanowicz and colleagues (2009) measured the same tension in ecological flow networks. Networks honed too tightly to one environment become "brittle" and collapse under minor novel disturbances; networks with too little organization lack what they need to survive. The ones that endure sit in a "window of vitality" between the two.

What we look for is a habit of judging a design across the range of conditions it will meet instead of at the expected one. Two moves give it away. The candidate asks which fluctuation would break the system, and checks whether a backup fails independently of what it backs up. Two backup generators fed by the same fuel line are one generator on the day the fuel stops.

Robustness costs money, and neither Kitano nor Ulanowicz treats it as something to maximize. A candidate who wants slack everywhere has missed the trade-off as surely as one who wants none. We also know of no cognitive study that measures this reasoning the way Cronin and colleagues measured accumulation. So the room task on robustness is scored and shown, and it stays out of the decision rule until we have data on it.

Existing tests, and what we kept

No established test measures Meadows' practice directly. We studied how the nearby instruments are built, then left them in the lab.

• Stock-and-flow items in the Cronin, Gonzalez, and Sterman line are the model for the inbox and bench items. On the phone the item is spoken, and in the room it comes as a table. The error shows up either way.
• Scenario tasks such as Grohs and colleagues' system-thinking prompt, and discovery tasks in the MicroDYN tradition, show why it is worth having people perform a model instead of rating themselves. We wrote short original items, borrowing none of those instruments, so the screen can be given twice in an hour.
• Raven-style matrices, complex span, the Wisconsin Card Sorting Test, and mental rotation measure relational induction, workspace, rule change, and spatial skill. They make good research covariates, but they shouldn't drive a screening decision: a person can pass all of them and still draw the stock as a copy of the flow.

References

1. Meadows, D. H. (2008). Thinking in Systems: A Primer. Chelsea Green.
2. Meadows, D. H. Dancing with Systems. Donella Meadows Project.
3. Meadows, D. H. Leverage Points: Places to Intervene in a System.
4. Cronin, M. A., Gonzalez, C., & Sterman, J. D. (2009). Why don't well-educated adults understand accumulation? Organizational Behavior and Human Decision Processes, 108(1), 116–130.
5. Sterman, J. D. (1989). Modeling managerial behavior: Misperceptions of feedback in a dynamic decision making experiment. Management Science, 35(3), 321–339. doi:10.1287/mnsc.35.3.321
6. Halford, G. S., Wilson, W. H., & Phillips, S. (1998). Processing capacity defined by relational complexity. Behavioral and Brain Sciences, 21(6), 803–831. PMID 10191879
7. Kane, M. J., & Engle, R. W. (2002). The role of prefrontal cortex in working-memory capacity, executive attention, and general fluid intelligence. Psychonomic Bulletin & Review, 9(4), 637–671.
8. Chi, M. T. H., Feltovich, P. J., & Glaser, R. (1981). Categorization and representation of physics problems by experts and novices. Cognitive Science, 5(2), 121–152. doi:10.1207/s15516709cog0502_2
9. Holyoak, K. J., & Monti, M. M. (2021). Relational integration in the human brain. Journal of Cognitive Neuroscience, 33(3), 341–356. doi:10.1162/jocn_a_01619
10. Hobeika, L., and colleagues (2016). General and specialized brain correlates for analogical reasoning: A meta-analysis of functional imaging studies. Human Brain Mapping, 37(5), 1953–1969. doi:10.1002/hbm.23149
11. Gonzalez, C., Fakhari, P., & Busemeyer, J. R. (2017). Dynamic decision making: Learning processes and new research directions. Human Factors. doi:10.1177/0018720817710347
12. Fleming, S. M. (2026). Towards an integrative neuroscience of metacognition. Nature Reviews Neuroscience. doi:10.1038/s41583-026-01081-x
13. Hamant, O. (2024). Debunking the idea of biological optimisation: Quantitative biology to the rescue. Quantitative Plant Biology, 5, e3. doi:10.1017/qpb.2024.3 (open access, PMC11016357)
14. Hamant, O. (2022). La troisième voie du vivant. Odile Jacob.
15. Hamant, O. (2023). Antidote au culte de la performance. La robustesse du vivant. Gallimard, Tracts n° 50. Presented by ENS de Lyon.
16. Hamant, O. (2024). En basculant de la performance à la robustesse, nous entrerons dans un monde viable. Interview, Millénaire 3, Métropole de Lyon.
17. Kitano, H. (2004). Biological robustness. Nature Reviews Genetics, 5(11), 826–837. doi:10.1038/nrg1471
18. Ulanowicz, R. E., Goerner, S. J., Lietaer, B., & Gomez, R. (2009). Quantifying sustainability: Resilience, efficiency and the return of information theory. Ecological Complexity, 6(1), 27–36. doi:10.1016/j.ecocom.2008.10.005 (author's PDF)
19. Lalani, B., Gray, S., & Mitra-Ganguli, T. (2023). Systems thinking in an era of climate change. Frontiers in Integrative Neuroscience, 17, 1145744. doi:10.3389/fnint.2023.1145744

The long note marks its neural claims as indirect. The construct has a literature, but nobody ran the Meadows tasks as the experiment. Keep them marked that way here.

Continue to the testing approach