SciencePediaIn our daily lives, we intuitively understand the world as local: an object's behavior is dictated by its immediate surroundings. This principle of locality is a cornerstone of classical physics and many successful scientific theories. However, a deeper look at nature reveals a universe that is far more interconnected, where what happens here can be fundamentally dependent on what is happening over there. These phenomena, known as nonlocal effects, represent a significant departure from our classical intuition and expose a critical knowledge gap in purely local descriptions of reality. This article bridges that gap by providing a comprehensive exploration of nonlocality. In the first chapter, "Principles and Mechanisms," we will investigate the theoretical and physical origins of nonlocal effects, from quantum mechanical correlations to the structural mechanics of materials, and examine why local models falter. Following that, the "Applications and Interdisciplinary Connections" chapter will take us on a journey across scientific disciplines to witness how nonlocality is not an esoteric concept but a vital principle for understanding everything from nanomaterials to the structure of the cosmos. By delving into these interconnected concepts, we will uncover a more complete and nuanced picture of the physical world.
In our everyday experience, and in much of classical physics, we are accustomed to a local world. The motion of a billiard ball is determined by the forces acting on it at this very moment, right where it is. The temperature at one point in a room is influenced by the temperature of the air molecules immediately adjacent to it. This seemingly obvious principle, that an object is only directly influenced by its immediate surroundings, is the bedrock of theories built on differential equations. The rate of change of a quantity at a point in space and time depends only on the properties of the field—its value, its gradient, its curvature—at that very same point.
For a long time, this local worldview was spectacularly successful. It even has a profound justification in the quantum world through what the great physicist and chemist Walter Kohn called the Principle of Near-Sightedness. In many materials, particularly metals where charges can move to screen out disturbances, or in insulators where electrons are "gapped" and it costs a lot of energy to excite them, the electronic system is remarkably provincial. A change in the electric potential at one location has an effect that dies off rapidly with distance. The electron density at a given point is largely determined by its local environment. This principle explains why relatively simple, local approximations in our most powerful theories often work surprisingly well. In a similar vein, when we study the mechanics of a large structure, we can often average over the complex, microscopic arrangements of atoms and treat the material as a smooth, continuous medium whose properties at a point are determined locally, a process known as homogenization.
But nature, in her full subtlety, is not always so near-sighted. There are crucial phenomena where what happens here depends fundamentally on what is happening over there, far away. These are nonlocal effects, and understanding them opens a door to a deeper and more accurate picture of the world, from the behavior of nanomaterials to the chemical bonds that make up life.
The first signs that a purely local view is incomplete often appear as small but persistent failures of our best local theories. Imagine trying to describe the gentle attraction between two neutral atoms, like an argon atom floating above a sheet of graphene. This is the realm of van der Waals forces (or more specifically, London dispersion forces), the subtle tugs that hold liquids and solids together, allow geckos to climb walls, and play a crucial role in the folding of proteins.
Where does this force come from? Even in a neutral atom, the electron cloud is not static; it is a shimmering, fluctuating quantum entity. For a fleeting instant, the electrons might be slightly more on one side than the other, creating a tiny, instantaneous dipole. This dipole creates an electric field that then influences the electron cloud of a nearby atom, polarizing it in sympathy. The result is a synchronized dance of fluctuating dipoles, leading to a weak, net attractive force. This attraction is a correlation effect: the motion of electrons in one atom is correlated with the motion of electrons in the other, even across empty space.
A local theory can't see this. A theoretical tool like the Local Density Approximation (LDA) in Density Functional Theory (DFT) tries to calculate the energy of the system by looking only at the electron density at each single point in space. From the perspective of a point in the vacuum between the atom and the graphene, the density is zero, so it contributes nothing to the interaction energy. The theory is blind to the long-range correlated dance of electrons and therefore completely misses the van der Waals attraction. This is a profound failure, a direct consequence of forcing a nonlocal phenomenon into a local straitjacket.
An even deeper form of nonlocality is woven into the very fabric of quantum mechanics by the Pauli exclusion principle. This principle states that no two identical fermions (like electrons) can occupy the same quantum state. A more dramatic way to put it is that every electron in the universe is fundamentally aware of every other electron. The total wavefunction of a system of electrons must be antisymmetric, meaning if you swap any two electrons, the sign of the wavefunction flips. This requirement creates what is called an exchange interaction, which is not a force in the classical sense, but a purely quantum statistical effect that profoundly influences energy. The exchange energy is inherently nonlocal; calculating it requires knowing where all the electrons are, not just what the density is at a single point.
Local approximations in DFT struggle mightily with this. They try to approximate this nonlocal exchange effect with a local recipe, which leads to famous pathologies like the self-interaction error: because the local approximation can't properly enforce the exclusion principle, an electron can spuriously interact with itself. Imagine a pandemic model where people are not immune to catching a virus from themselves; it's a similarly unphysical situation. This error, in turn, can lead to other failures, like the incorrect prediction that electrons will unnaturally spread out over multiple molecules. In the extreme world of theoretical physics, there's a fascinating insight: in a hypothetical universe with infinite spatial dimensions and purely local, on-site interactions, the quantum world suddenly becomes local again. The mathematical description of electron correlations, the self-energy, collapses into a purely single-site object. This is the very trick that makes a powerful method called Dynamical Mean-Field Theory (DMFT) work. But as soon as you introduce interactions that connect different sites, nonlocality comes roaring back, and the simple picture breaks down.
If the world is sometimes nonlocal, how do we build this property into our theories? There isn't one single answer; rather, there is a rich collection of strategies, each tailored to the physical origin of the nonlocality.
In the world of materials, especially at the nanoscale, the idea that the state of matter at a point depends on its neighbors becomes tangible. In nonlocal elasticity, pioneered by A. Cemal Eringen, this is formalized by saying that the stress (the internal forces) at a point is a weighted average of the elastic response over a finite neighborhood. Instead of the simple local Hooke's law, , where is the strain, the nonlocal version becomes an integral:
Here, is a kernel function that weighs the influence of the strain at neighboring points . This seemingly small change has dramatic consequences. It predicts that the stiffness of a material can depend on its size or the shape of its deformation. For example, a nanobeam subjected to a short-wavelength bending deformation will appear softer than predicted by classical theory, because the averaging process smooths out the sharpest parts of the strain field. This is a real, experimentally observed size effect.
Other situations call for different approaches. Near a sharp crack tip or under a nanoindenter, the strain can change so abruptly that the material resists not just the strain itself, but the gradient of the strain. This leads to strain gradient elasticity, which adds higher-order derivatives to the energy. This theory is particularly good at explaining the observation that smaller things are often stronger, because it effectively models the pile-up of microscopic defects known as dislocations. For specially engineered materials like chiral metamaterials, whose internal structures can rotate independently of the surrounding material, an even more exotic theory called micropolar elasticity is needed. It introduces an independent field for these microrotations, leading to new types of stresses called couple-stresses. This beautiful plurality of theories shows that "nonlocal" is not a monolithic concept, but a rich language for describing the diverse ways in which an object's state can depend on more than just a single point.
In quantum chemistry, the quest for including nonlocality has led to a beautiful hierarchy of theories, famously dubbed Jacob's Ladder by John Perdew. Each rung on the ladder adds a new ingredient to the DFT functional, bringing it closer to the exact (and fully nonlocal) truth.
Rung 1 (LDA) & Rung 2 (GGA): These are the semi-local rungs we've visited, based on the density and its gradient . They are near-sighted by construction.
Rung 3 (meta-GGA): This rung adds the kinetic energy density, which provides some implicit information about the local electronic structure. It can, for instance, recognize a region with only one electron and treat it more accurately, but it remains a semi-local theory.
Rung 4 (Hybrid Functionals): This is the great leap to nonlocality. These functionals mix in a fraction of the truly nonlocal exact exchange energy. This dose of nonlocality provides a partial antidote to the self-interaction error that plagues lower rungs. It dramatically improves the prediction of properties sensitive to long-range effects, like the energy of chemical reactions and the behavior of an electron stripped far away from its parent molecule.
Rung 5 (Double Hybrids): To finally capture the nonlocal correlation of van der Waals forces, we must climb to the fifth rung. Here, we add another nonlocal ingredient, this time a correlation term borrowed from traditional quantum chemistry methods (like second-order Møller-Plesset theory, MP2, or the Random Phase Approximation, RPA). This term is explicitly built to describe the correlated fluctuations of electrons over long distances.
Clever modern strategies, known as range-separated hybrids, offer a pragmatic solution. They use a computationally cheap local model for short-range interactions where they work well, and switch to an expensive but accurate nonlocal model for the long-range part where nonlocality is essential. It's a beautiful piece of theoretical engineering that gives us the best of both worlds.
Embracing nonlocality solves many old problems, but it introduces new, subtle challenges. The mathematical elegance of local, differential equations is replaced by the more cumbersome world of integral or higher-order equations. A particularly thorny issue arises in nonlocal material models: what happens at the boundary?.
If the stress at a point is an average over a neighborhood, what happens to a point near the surface of the object? Part of its neighborhood is missing—it is outside the material. This truncation of the integral means the material appears artificially soft at its surfaces. This is not just a numerical annoyance; it's a fundamental paradox that can violate conservation laws if not handled carefully.
Physicists and engineers have devised clever remedies. One approach is to renormalize the averaging kernel near the boundary, essentially forcing it to be properly normalized. But this trick often comes at a steep price: it can break the mathematical symmetry of the model, which means that a simple, elegant expression for the material's stored energy may no longer exist. An alternative is to imagine a fictitious layer of material extending just beyond the real boundary. This allows the integral to be completed, preserving the model's symmetry, but requires solving an additional problem to determine the properties of this imaginary layer. The fact that so much thought must go into simply defining a nonlocal model on a finite object highlights that this is a vibrant and active frontier of research. The journey from a simple, local world to a complex, nonlocal one is a profound shift, revealing a universe that is far more interconnected than we might first imagine.
In the last chapter, we took a careful look at the principle of locality—the simple, intuitive idea that an object is only directly influenced by its immediate surroundings. We saw that this is often a wonderfully useful approximation, but an approximation nonetheless. The universe, it turns out, is full of situations where what happens here is profoundly tied to what’s happening over there. This “nonlocality” isn’t some esoteric footnote; it’s a crucial feature of reality that emerges in a dazzling array of fields. Now, let’s go on a journey to see this principle in action, from the microscopic machinery of life to the grand structures of the cosmos. We will find that appreciating nonlocality is not just about correcting our old models, but about discovering entirely new phenomena.
Our everyday experience is with objects large enough that we can safely ignore the jostling of individual atoms. A steel beam is a steel beam, and if we make it half as big, we expect it to behave in roughly the same way, just scaled down. But what happens when you shrink things so much that the "external" size of the object starts to become comparable to the "internal" reach of the forces between its atoms? At this point, the local approximation breaks down spectacularly.
Consider the challenge of building machines at the nanoscale. Imagine you want to use a tiny nanowire as a structural element. How stiff will it be? If you use the classical equations of continuum mechanics, which are strictly local, you’ll get an answer. But it will likely be the wrong one. An atom in the wire doesn't just feel the pull of its immediate neighbors; it interacts with a small cloud of atoms around it. In a large beam, this interaction range is negligible. But in a nanowire just a few dozen atoms thick, that range might be a significant fraction of the wire's diameter. This means the stress at one point depends on the strain in a whole neighborhood. This nonlocal effect fundamentally changes the wire’s bending stiffness. The key parameter telling us when to worry is the ratio of the internal interaction length, let's call it , to the diameter of the wire, . When the ratio is no longer vanishingly small, nonlocal corrections become essential for accurate engineering.
This same story unfolds in the realm of nanophotonics, where we manipulate light with metallic nanoparticles. According to local electromagnetic theory, the color of light scattered by a tiny gold sphere depends on the collective oscillation of its electrons—a surface plasmon—but this color shouldn't depend on the sphere's size. Yet, in the lab, we see a beautiful blue-shift as the particles get smaller. Why? The "sea" of free electrons within the metal acts like a fluid. A local model treats this fluid as infinitely compressible, but in reality, it has a kind of pressure. When you try to squeeze the electron sea by confining it in a smaller and smaller sphere, this pressure pushes back. This nonlocal hydrodynamic effect means the electron oscillations are stiffer in smaller particles, resonate at a higher frequency, and thus scatter bluer light. The particle's radius becomes a critical parameter, a fact lost on any local theory.
In both the mechanical wire and the optical particle, the lesson is the same: when you enter the nano-world, you can no longer assume that an object's properties at a point are determined solely by the fields at that same point. The very geometry of the object acts as a global constraint on its internal workings.
Nonlocality isn’t just about being very small. It’s about the pattern of connections. Sometimes, long-range links are so powerful that they completely override the tendencies of local interactions. Nature offers a stunning example in the machinery of life itself: protein folding.
Imagine trying to predict the three-dimensional shape of a peptide, a small protein, from its sequence of amino acids. A common computational approach is to slide a window along the sequence and, based on the local pattern of amino acids, predict if the central one is part of a helix, a sheet, or a disordered coil. This is a purely local strategy. When applied to certain venom peptides, like conotoxins, it fails miserably, predicting an almost entirely disordered, floppy chain. Yet, experiments reveal these molecules have a highly compact and rigid structure. The secret lies in a few key cysteine residues. These residues are often far apart in the linear sequence, but they are "instructed" by the laws of chemistry to form strong covalent disulfide bonds. These bonds act like powerful clamps, pulling distant parts of the chain together and forcing the entire peptide into a specific, intricate fold. The local propensities of the amino acids are irrelevant; the structure is dictated by these dominant, nonlocal connections. The prediction algorithm failed because it was blind to the global architectural plan.
A similar principle governs the behavior of things as mundane as a pile of sand. Dense granular materials don't flow like water. Their ability to flow—their "fluidity"—is not a simple local property. In a silo, a region of grains can become "jammed" due to stress, and this jammed state can create a rigid arch that supports all the material above it, profoundly influencing the flow far below. Sophisticated models of granular flow discard local rheology and instead propose that the fluidity at a given point depends on the gradients of stress in a surrounding volume. The tendency to flow is a cooperative, nonlocal phenomenon, where a region's fate is tied to the state of its entire neighborhood.
Of course, the most famous arena for nonlocality is the quantum world. In a quantum system, particles can become entangled, their fates intertwined no matter how far apart they are. But this "spookiness" is not just a philosophical curiosity; it has tangible, measurable consequences and is even becoming an engineering resource.
Consider a superconductor. What makes it "super" is that its electrons pair up and condense into a single, macroscopic quantum state that spans the entire material. You can no longer think of the electrons as individuals. Now, try to probe this material with a magnetic field. In an ordinary metal, the resulting current at a point depends on the electric field at that same point . This is Ohm's law—a local law. But in a clean superconductor, the current's response is nonlocal. The coherent cloud of electron pairs responds as one. The current at depends on the electromagnetic field averaged over a region defined by the size of the electron pairs (the coherence length). This fundamental nonlocality leads to concrete, testable predictions. For instance, in certain unconventional superconductors, it alters the very nature of how the magnetic penetration depth changes with temperature, causing the dependence to switch from being proportional to to proportional to at very low temperatures. It also introduces subtle corrections to fundamental properties like the lower critical magnetic field, the point at which flux vortices first enter the material.
Perhaps most excitingly, physicists are now trying to harness nonlocality to build a quantum computer. One promising approach, measurement-based quantum computing, requires preparing a vast, highly entangled "cluster state" as a resource. How can you create such a globally correlated state in a finite amount of time? If you only have local, short-range interactions in your system—like atoms only bumping into their nearest neighbors—information can only spread at a finite speed, limited by a kind of "light cone." A quick, global operation (a "quench") would fail to establish the necessary long-range order. The solution? Engineer a system whose Hamiltonian contains nonlocal, long-range interactions. Theory shows there is a critical threshold for how quickly these interactions must fade with distance. If they decay too fast (e.g., faster than in two dimensions), the system behaves locally. If they decay slowly enough, the system enters a nonlocal regime where a single operation can, in principle, create the global entanglement needed for a universal quantum computer. Nonlocality is no longer just a feature to be described; it's a design specification.
The theme of nonlocality scales up to the level of engineering design and even to the structure of stars.
In modern engineering, we often use computers to find the "optimal" shape for a bridge or an airplane wing—the lightest design for a given strength. The objective, overall stiffness, is an inherently global property. It depends on how the entire structure deforms under a load. Mathematically, this is a nonlocal problem. The computer model, however, breaks the structure down into a grid of simple, local pieces called finite elements. And here, a fascinating pathology can arise. A naive optimization algorithm, with its global, nonlocal view, is clever enough to discover and exploit weaknesses in the simple local elements. It may produce bizarre "checkerboard" patterns of material and void, which the local elements incorrectly interpret as being extremely stiff. The resulting design is a numerical artifact, useless in the real world, born from the conflict between a nonlocal objective and a flawed local model. Understanding this interplay is essential for creating robust design tools.
Finally, we journey to the cosmos. The standard model of stellar structure, the Lane-Emden equation, is built on local physics: the pressure at a point depends only on the density and temperature at that point. But what if the "stuff" making up a star—perhaps some form of exotic dark matter—obeys nonlocal laws? Theoretical astrophysicists explore this very question by adding nonlocal terms directly into the star's equation of state, often using the mathematical machinery of fractional calculus. This modifies the classic stellar structure equations, opening the door to entirely new types of cosmic objects with different density profiles and stability properties than ordinary stars. Here, nonlocality serves as a powerful tool for imagination, allowing us to ask "what if?" and explore possibilities for the universe's structure beyond what our local intuition suggests.
From the stiffness of a nanowire to the folding of a protein, from the color of a gold nanoparticle to the quest for a quantum computer, a single, unifying theme emerges. Our simple, local picture of the world is a powerful and often sufficient approximation. But it is not the whole story. Whenever a system is governed by global constraints, long-range physical connections, or collective quantum coherence, the "whole" becomes profoundly more than the sum of its local parts. Understanding the principle of nonlocality is to see the invisible threads that weave the fabric of our universe together, connecting the near and the far, the small and the large, in ways that are at once subtle, beautiful, and deeply fundamental.