Aggregatism: See the Forest, Then the Trees

A forest is more than just an assembly of trees. This simple observation, almost self-evident in daily life, points to a profound intellectual problem: how should we understand the relation between parts and wholes? Trees constitute the forest, yet the forest is not merely a numerical sum of trees. It has its own ecology, dynamics, and functions. The forest, for example, stores carbon across landscapes, regulates water through soils, harbors biodiversity, and sustains resilience in ways no single tree can. To stare only at the tree is to risk missing the forest altogether. The familiar phrase “can't see the forest for the trees” thus expresses not merely a perceptual failure, but a methodological warning. The relation between part and whole—mereology—has long occupied philosophy. Holism asks us to attend to integration, organization, and totality; reductionism directs our gaze downward, toward constituent units and their intrinsic properties. Modern science has overwhelmingly favored the latter. Reductionism maintains that the properties of the whole are reducible to, and predictable from, the properties of its parts: “the whole has no properties beyond those of its component parts” 1. This approach has been extraordinarily successful. From Democritus's early proposal to the experimental programs of modern chemistry, physics, and biology, reductionism has furnished science with clarity, rigor, and explanatory power. So influential has it become that, in practice, scientific inquiry is often assumed to be reductionist inquiry. Nowhere is this orientation more evident than in molecular science. If the material world is composed of molecules, and if a molecule is, as Merriam-Webster puts it, “the smallest particle of a substance that retains all the properties of the substance” 2, then it seems natural to suppose that full understanding of materials must begin with molecular species. From this assumption arises molecularism: the conviction that the key to understanding matter, and even life itself, lies in the elucidation of molecular parts 3. Molecularism has become one of the deepest working assumptions of laboratory science. It is not merely a theory of matter, but a theory of inquiry: find the molecule, determine its structure, identify its property, and then proceed outward toward application. This logic has undeniable force. A molecularist study usually begins with a single molecular species (Figure 1a), either synthesized de novo or selected from existing compounds, including natural products. When this molecule displays a desired property—represented here as “1” in Figure 1—the next step is to examine its aggregate, since practical use almost always concerns aggregates rather than isolated molecules. If the aggregate preserves the molecular property, the material can then be optimized for application. This pathway, denoted M1,1,1 (where M refers to “molecule”), is orderly, efficient, and often highly productive. It is one reason molecularism has become the paradigm of what Kuhn called “normal science” 4. Yet, no paradigm is universal. Molecularism is powerful, but it is not omnipotent. It struggles particularly with systems in which properties disappear or emerge upon aggregation. In an M1,0,0 system, a molecule possesses the desired property, but the aggregate does not. In such cases, success at the molecular level turns into failure at the material level. The property that motivated the study vanishes precisely where practical relevance begins. Aggregation-caused quenching offers a classic example: a molecular species emits light in isolation yet loses emission in the aggregated state. What seemed promising at the molecular scale becomes hopeless or useless for a particular application (e.g., fabrication of an OLED device) once the aggregate is formed. Similar stories can be told in pharmaceutical research, where lead compounds selected from vast screening campaigns may lose efficacy when converted into administrable forms. The cost is not merely theoretical disappointment, but wasted time, effort, and resources. Even more consequential is the M0,1,1 system. Here, the molecule itself lacks the target property, and a molecularist workflow may, therefore, terminate the project at once. The underlying assumption is simple and rarely questioned: if the part does not possess the property, the whole will not either 1. But this assumption, though intuitively appealing, can be profoundly misleading. It discourages us from looking for aggregate-level phenomena absent in the constituent molecules and thus blinds us to forms of functionality that arise only through collective organization. Aggregation-induced emission (AIE) is the paradigmatic counterexample 5. A molecule that is nonluminescent in isolation becomes emissive upon aggregation. What does not exist at the level of the part appears at the level of the whole. This is not a trivial exception. It is a philosophical challenge. If new properties can arise at the aggregate level—properties neither visible in nor straightforwardly inferable from isolated molecules—then the aggregate cannot be treated merely as a passive container of molecular traits. It must be regarded as a level of reality worthy of direct scientific attention. After all, isolated molecules are largely invisible to us and rarely directly usable by themselves. Aggregates are what we see, touch, handle, process, and rely upon in the material world. For this reason, we propose aggregatism as a research paradigm for aggregate science 6, one that shifts both the starting point and the center of gravity from molecular species to aggregate systems (Figure 1b). Aggregatism does not reject molecularism; rather, it reorders scientific attention. It asks us to begin from the level at which materials actually exist and function. In an A1,1,1 system (where A denotes “aggregate”), this approach readily accommodates cases in which aggregate behavior corresponds to molecular expectation. But its true value appears in disappearance and emergence systems. In an A1,0,0 system, aggregatism allows earlier recognition of practical failure, avoiding excessive investment in molecules whose attractive properties do not survive aggregation. In an A0,1,1 system, by contrast, it keeps inquiry alive where molecularism would likely abandon it. The aggregate can be explored directly for function and application, even before its emergent behavior is fully explained. The scientific challenge then shifts: no longer “Why did the molecule fail?” but “How does the whole become more than the sum of its parts?” At this point, the discussion enters the domain of emergentism. Molecularism is a mature methodology, institutionally entrenched and procedurally refined. Aggregatism is younger, less codified, and, therefore, more open-ended. That very openness is part of its intellectual appeal. Emergence, after all, is among the most intriguing ideas in both science and philosophy. A property is emergent when the whole displays something absent from its individual parts 7. In this sense, AIE is not only a photophysical phenomenon but also an epistemological lesson. Nature, and especially life, is filled with such lessons. The world is replete with systems in which organization matters, interactions matter, context matters, and novelty appears not despite complexity but because of it. Emergent phenomena have often been regarded as difficult, if not impossible, to predict from lower-level description alone, in part because complex systems are often nonlinear. This makes emergence scientifically elusive and philosophically provocative. A bottom-up route from molecular parts may eventually uncover such properties, but often only through serendipity. A top-down route—beginning from aggregate behavior and then tracing downward to its organizational basis—may offer a more practical way forward. It was precisely this conviction that motivated our systematic investigation of AIE 6. What first appeared as an abnormality in the part–whole relationship became, on closer inspection, an entry point into a more general mode of understanding. That success encouraged us to venture further. We subsequently explored a range of emergent systems, including room-temperature phosphorescence, circularly polarized luminescence, mechanoluminescence, and clusteroluminescence (CL). CL is especially illuminating. Some common plastics and biomasses—polyesters, polyamides, celluloses, alginates, and so on—can become photoluminescent as clusters in the solid state 8. Yet, at the molecular level, these materials are π-electronically nonconjugated and, according to standard photophysical expectations, should not emit. Their isolated molecular species in dilute solutions are indeed nonluminescent. By molecularist criteria, they would appear unpromising. And yet, in the aggregate, luminescence emerges. This is precisely the sort of phenomenon that invites a philosophical pause. The molecular parts are not irrelevant; they remain indispensable. But their significance lies not simply in what they are individually, but in how they interact, arrange, and cohere. Through systematic investigation, we concluded that through-space interactions and the organizational arrangement of the molecular parts underlie CL. The lesson is both scientific and conceptual: the source of function may reside not in isolated entities but in relations, proximities, and patterns of assembly. Once this is appreciated, emergence appears less as an anomaly and more as a natural consequence of organized matter. Aggregatism, then, is more than a technical proposal. It is a change in scientific perspective. It invites us to look first at the whole and only then at the parts—not because parts are unimportant, but because wholes may possess a kind of explanatory primacy in systems where organization generates novelty. The forest should not be treated as an afterthought to the tree. In many cases, the forest is where the real questions begin. This perspective may be especially consequential for the life sciences. One may know a great deal about DNA, RNA, proteins, and carbohydrates and still remain far from understanding life itself 1, 9. These biomacromolecules, considered in isolation, are not alive. Life emerges through collective interactions, spatial organization, dynamic regulation, and continual exchange with the environment. To explain life solely by cataloging its molecular components is like trying to explain a forest by listing the anatomy of trees. Necessary, yes; sufficient, no. For this reason, we suggest a practically workable two-way framework for aggregatist research, particularly for emergent systems. One begins with the aggregate, because that is where function appears and application becomes possible. From there, one moves upward toward deployment of the aggregate as a useful material, while also digging downward to uncover the interactions, arrangements, and organizational principles among molecular components (Figure 1b). In this way, application and mechanism advance together, neither postponed to the other. We believe that aggregatism can open fertile ground for scientific imagination. It may generate new models, new hypotheses, and new theories for understanding natural processes that resist explanation within a purely molecularist framework. It may also provide a conceptual basis for the rational design of biomimetic emergent systems and advanced functional materials. More broadly, it reminds us that science progresses not only by seeing deeper into the small, but also by learning how to think more clearly about the large, the organized, and the collective. To understand nature, one must sometimes resist the “instinct” to begin with the smallest unit. Sometimes the right place to start is the world as it appears: structured, relational, and already assembled. Sometimes, in other words, one must see the forest first. Author thanks the financial support from the Major Program of the National Natural Science Foundation of China (22595400 and 22595401) and the Guangdong Basic Research Center of Excellence for Aggregate Science. The author declares no conflicts of interest.