The scale you can’t see: rethinking the “time problem” through Levin’s ecology

This week, in a meeting about Research Kernel Insights, the conversation tilted unexpectedly into ecology — and into a question I have not stopped thinking about since: am I observing my research life at the wrong scale? The reference that anchored the discussion was Simon Levin’s 1992 paper The Problem of Pattern and Scale in Ecology (Levin, 1992), and what follows is my attempt to translate that idea into something practical for researchers.

Levin’s point: scale is not a detail, it is part of the system

Levin’s central claim is that there is no single, natural scale at which an ecological system should be described. The pattern you see depends on the scale of observation, and the mechanisms that generate a pattern often live at a different scale than the one where the pattern is observed (Levin, 1992; Chave, 2013). Wiens (1989) operationalises this by introducing the concepts of grain (the resolution of any given observation) and extent (the total scope covered) — a reminder that every measurement is already a scale choice, not a neutral act.

His favourite illustration is plankton and krill in the ocean. At large scales (kilometres, days), patchiness is sculpted by physics — advection, eddies, upwelling. Zoom in (metres, minutes), and the same patches are shaped by biology — behaviour, predation, reproduction (Levin, 1992; Wiens, 1989). Same ocean. Same organisms. Specific changes in process as you move across scales.

“A system is not chaotic just because it looks messy at one scale. It is usually a sign that you are reading it at the wrong resolution, or with the wrong vocabulary.

Levin’s first move, in fact, is to insist that scale, pattern and process must be defined operationally before any modelling begins (Levin, 1992). No definition, no chain of reasoning.

From the ocean to the researcher’s life

The same multi-scale logic applies, almost line by line, to a research career. Each phase — Bachelor’s, Master’s, PhD, Postdoc, Assistant/Associate Professor, Full Professor, Industry Researcher/Scientist, and Industry R&D / Principal Scientist — is a different “scale” with its own dominant process, its own observable pattern of progress, and its own critical vocabulary that has to be defined before you can act on it.

Each row is a different ocean. Each row has its own krill and its own eddies. Importing the metrics of one row into another — judging a PhD week with industry-sprint expectations, or evaluating a professor’s month with Master’s-project deliverables — is exactly the mistake Levin warned ecologists about: lifting a pattern from the wrong scale and treating it as universal truth (Levin, 1992).

Holling and Gunderson (2002) formalise a complementary insight through panarchy: nested adaptive cycles where smaller, faster levels experiment and larger, slower levels conserve accumulated memory — the PhD as rapid exploration, the tenured professorship as conservation and consolidation.

The unified processes — curiosity, asking, writing, collaborating, caring — recur at every level. What changes is the expression: the cadence, the unit of output, the kind of attention required. Recognising the recurrence is what allows a researcher to act coherently across years without losing the thread.

From ecology to research life: clustering as the bridge

One reader pointed out, fairly, that there is a gap between Levin’s ecology and the career table above — an unstated step where the analogy actually does its work. The gap was closed, quietly, long before Levin wrote his 1992 paper. Herbert Simon (1962) showed that complex systems — in physics, biology, and social organisation alike — share a common structural property: they are nearly decomposable hierarchies, with strong internal coupling within each level and weak coupling between levels (Simon, 1962). This is the architectural principle that makes scale-separation possible at all.

Levin himself bridges the gap explicitly in later writing: Chave and Levin (2003) extend the pattern-and-scale framework to socioeconomic systems, treated as complex adaptive systems whose macroscopic patterns emerge from local interactions among heterogeneous agents. They invoke Simon and Ando’s (1961) near-decomposability to justify analysing careers — or any multi-scale system — as clusters of activities with strong internal coupling on fast time scales and weak external coupling on slow ones (Chave & Levin, 2003).

The practical benefit of clustering by career scale is therefore not productivity folklore. It is the canonical move for any multi-scale complex adaptive system: it reduces dimensionality, makes the fast/slow dynamics tractable, and prevents the cross-scale confusion that Levin (1992) warned against. Once tasks are grouped by their native scale, the system becomes legible.

From “I don’t have time” to “I have a system to read”

The complaint “I don’t have time” treats time as a flat, scalar resource — a bucket that empties. Levin’s lens treats time as structured, multi-scalar, and process-dependent (Levin, 1992). The minute you adopt that lens, the complaint mutates: the question is no longer “how do I get more time?” but “which scale am I operating at right now, and is the process I’m trying to run native to that scale?”

Building a well-defined chain of reasoning across these scales is what allows you to step back from the “problem” of time and start treating it as a system. The practice has three steps:

• Define the critical words of your current phase before you plan anything (what exactly counts as progress this year?).
• Locate the scale of every task you carry — is this an advection task (slow, structural, large) or a behaviour task (fast, local, reactive)? Newport (2016) draws the same boundary when he distinguishes deep work from shallow work: cognitively demanding tasks belong to slow scales and must be protected from fast-scale interruptions.
• Chain the reasoning by showing how unified processes — learning, writing, asking, collaborating — reappear at each scale, so you stop reinventing them and start recognising them (Levin, 1992; Chave, 2013).

When that chain is in place, “lack of time” stops being a feeling and becomes diagnosable. You can point at the row in the table where the friction lives, name the process that is misaligned with its scale, and adjust — exactly the way an ecologist would isolate a mechanism inside a patch.

“The ocean is not chaotic. We are just standing too close, or too far. Same for a research career.

References

• Chave, J. (2013). The problem of pattern and scale in ecology: what have we learned in 20 years? Ecology Letters, 16(s1), 4–16.
• Chave, J., & Levin, S. A. (2003). Scale and scaling in ecological and economic systems. Environmental and Resource Economics, 26(4), 527–557.
• Gunderson, L. H., & Holling, C. S. (Eds.). (2002). Panarchy: Understanding transformations in human and natural systems. Island Press.
• Levin, S. A. (1992). The problem of pattern and scale in ecology: The Robert H. MacArthur Award Lecture. Ecology, 73(6), 1943–1967.
• Newport, C. (2016). Deep work: Rules for focused success in a distracted world. Grand Central Publishing.
• Simon, H. A. (1962). The architecture of complexity. Proceedings of the American Philosophical Society, 106(6), 467–482.
• Simon, H. A., & Ando, A. (1961). Aggregation of variables in dynamic systems. Econometrica, 29(2), 111–138.
• Wiens, J. A. (1989). Spatial scaling in ecology. Functional Ecology, 3(4), 385–397.