Abstract. The ontology of physical science concerns the kinds of entities, structures, processes, and properties that a warranted interpretation of physical theories commits us to recognizing. The problem joins an ancient inquiry into nature with modern disputes about theoretical reference, experimental intervention, scientific realism, mathematical representation, classical field theory, spacetime, quantum mechanics, and quantum field theory. This monograph reconstructs the transition from Aristotelian substances and Newtonian bodies through the transformation of physical geometry and the emergence of atomic, field-theoretic, and quantum descriptions. It analyzes Quinean ontological commitment, constructive empiricism, entity realism, explanationist and structural realisms, theory underdetermination, the problem of historical discontinuity, and the relationship between models and target systems. It then tests competing ontologies against particles, fields, gauge redundancy, wave functions, quantum nonseparability, spacetime structure, composite physical systems, effective theories, emergence, and experimental evidence. The central result is a disciplined account of ontological inference: a mathematical variable acquires a warranted physical interpretation only through a specified representational role, empirically supported explanatory and interventionist relations, and robustness under defensible reformulations. This account preserves the reality of physical phenomena while distinguishing strength of evidence, dependence on theory, and the open metaphysical interpretation of successful formalism.
Keywords: physical ontology; scientific ontology; physical entities; scientific realism; theoretical entities; ontological commitment; electrons; fields; quantum field theory; spacetime; wave function; structural realism; entity realism; constructive empiricism; underdetermination; gauge symmetry; emergence; effective field theory; scientific representation; naturalized metaphysics.
The Philosophical Problem of Physical Being
Physical science makes claims about a world that exceeds ordinary perception. An electron leaves a trace, a detector registers an excitation, a gravitational wave changes relative distances, and a mathematical model assigns amplitudes to possible outcomes. Each episode raises an ontological question: what must there be for the representation and the experimentally established behavior to be as they are? The answer cannot be extracted merely by reading nouns out of scientific sentences. The same phenomenon can be represented by different mathematical formalisms; the same term can survive radical changes of theory; and mathematical structures may contain degrees of freedom whose differences have no empirical consequence. Scientific ontology is the philosophical investigation that orders these distinctions and asks which parts of successful scientific practice support existence claims.
The problem is acute because physical inquiry simultaneously extends knowledge and revises the categories in which knowledge is expressed. Classical mechanics described objects with determinate trajectories. Electromagnetism made fields indispensable to successful description. Relativity revised the status of space and time. Quantum theory complicated the assignment of determinate values and the individuation of particles. Quantum field theory unifies central calculations involving creation and annihilation with an ontology that philosophers continue to dispute. The existence of experimentally established phenomena is often secure while the best account of their fundamental bearers remains contested. Confusing these two achievements makes realism dogmatic and skepticism indiscriminate.
The organizing thesis of this study is that physical ontology demands three coordinated kinds of inquiry. Formal interpretation establishes what a theory says under an explicit semantics. Empirical evaluation determines which of its claims are supported by observation, independent detection, intervention, and explanatory success. Metaphysical analysis identifies the kinds and dependence relations that a defensible interpretation would posit. These inquiries constrain one another, although none alone supplies an exhaustive inventory of physical being. The resulting framework is both historically responsive and sufficiently exact to address foundational physics.
Definition, Scope, and Ontological Levels
A physical entity is, for the purposes of this research, an individual, field, process, system, event, or physically instantiated structure invoked as a constituent of the subject matter of physical science. The term "entity" is deliberately broad; its application alone does not certify that a theoretical posit exists. A scientific ontology is a reasoned account of the kinds and relations to which an accepted or defended scientific representation is ontologically committed. A physical ontology is the special case centered on the subject matter of physics and its explanatory connections to other material sciences. These are working definitions that specify what the inquiry examines and leave the success of particular ontological claims to argument.
Four distinctions prevent category mistakes. First, a target phenomenon is what an experiment or model seeks to investigate; its theoretical representation is a construct employed to describe or explain it. Second, a mathematical object belongs to a formal theory; a physical entity is what a physically interpreted theory purports to describe. Third, an ontological commitment of a formalization is a semantic consequence under its chosen interpretation; epistemic warrant is the justification for accepting that commitment as approximately true of reality. Fourth, fundamentality concerns relations of dependence and explanatory priority, whereas existence alone specifies that something is part of the world. A storm, a detector, and an atom may all exist without occupying the same level in a fundamental theory.
These distinctions also establish the monograph's boundaries. General questions about existence, universals, or metaphysical grounding are considered where they affect scientific ontology, while the fine structure of quantum dynamics, the complete history of cosmology, and the philosophical status of every special-science category remain independent research programs. The concern here is to explain how physical theories obtain justified ontological interpretation, what different kinds of entity claims amount to, and why even empirically triumphant science can underdetermine a unique metaphysical picture.
Nature and Physical Substance in Ancient Philosophy
The ancient Greek inquiry into physis framed nature as a domain of change requiring explanation. Aristotle's Physics distinguishes what exists by nature from artifacts in terms of an internal principle of motion and rest (Physics II.1, 192b, in the Hardie and Gaye translation). This approach makes the organization and developmental tendencies of a natural thing essential to explaining what it is. The material from which a living organism is composed cannot alone explain its life, while a merely geometrical description cannot establish what produces its characteristic activities. Aristotle's account of form and matter therefore embeds an ontology within a theory of explanatory dependence rather than treating the task as enumeration.
Ancient atomism supplied a markedly different proposal: the constitution and movements of bodies could be understood through indivisible constituents and void. In both the Aristotelian and atomist projects, ontology was inseparable from explanation, yet their explanatory ambitions differed. The former sought principles appropriate to different modes of change, and the latter sought a parsimonious underlying inventory from which visible diversity could arise. These are conceptual ancestors of disputes about whether today's physical ontology should center on individual microscopic units or on organized structures and processes. The analogy has limits: modern atomic physics does not vindicate the ancient atom as a physically indivisible solid particle, and present-day structural realism is not simply Aristotelian hylomorphism under a different name.
The historical lesson is methodological. Describing the most familiar objects does not settle what is physically basic, and postulating hidden constituents does not by itself establish their reality. Already in the classical discussion, competing ontologies were tested by the work they could do in explaining persistence, alteration, and the unity of natural systems. The subsequent scientific revolution transformed the methods available for adjudicating these proposals, while retaining the underlying question of how natural explanation bears on being.
Mechanics, Matter, and the Newtonian World
Newton's Philosophiae Naturalis Principia Mathematica (1687) joined mathematical laws of motion to a systematic account of forces and changing states. Its definitions and scholium distinguish absolute from relative measures of space, time, and motion. A particle-like body in this scheme can be modeled through position, mass, momentum, and forces, and large-scale motions become calculable through general principles. The intellectual transformation consisted not simply in naming material objects but in embedding their possible motions in a mathematically articulated explanatory architecture (Newton, 1687, Definitions and Scholium).
A central ontological tension follows. Newtonian space and time can be interpreted as structures existing independently of material occupants, while relational accounts treat spatial and temporal facts as constituted by relations among events or bodies. The formal success of mechanics does not immediately decide between these interpretations. Although Newton's own commitments extend beyond a merely instrumental account, separating the historical authorial view from the implications of a mathematical theory remains indispensable. Later debates over inertial motion, reference frames, and the status of background geometry show that the ontology of physical description is not exhausted by its inventory of moving masses.
Mechanics also established a powerful idealization practice. A point mass occupies a mathematically exact location while abstracting from spatial extension; perfectly rigid bodies and frictionless planes simplify dynamics; isolated systems make laws tractable. The success of such descriptions does not entail that the idealized object, exactly as represented, must exist. Rather, an idealization specifies the features that matter to an explanatory task. The distinction between a mathematically useful surrogate and a physically instantiated kind becomes unavoidable once idealized systems are used to infer properties of real systems.
Fields, Geometry, and the Transformation of Physical Description
Nineteenth-century electromagnetism made field descriptions central to physics. A field assigns quantities across a region, relating interactions through spatially distributed physical states instead of reducing every influence to the direct contact of discrete bodies. In Maxwellian theory, electromagnetic behavior is represented by equations whose physically significant features persist under changes in the way electromagnetic potentials are described. The field concept thus challenged the idea that the sole fundamental ontological units must be localized corpuscles, while simultaneously generating a new problem: which field variables genuinely describe physical degrees of freedom?
The transition from ether theories to electromagnetic field theories became a test case for philosophical realism. Worrall's reconstruction of the transition from Fresnel's optical theory to Maxwell's electromagnetism argues that preservation of mathematical relations across theory change can warrant a selective realist commitment to structure without preserving every earlier entity description (Worrall, 1989). The challenge to that argument is to specify precisely which structure survives, whether structurally similar equations have the same physical interpretation, and how to avoid treating formal continuity as ontological identity. Structural survival has philosophical force when its empirical and interpretative content are also understood.
Non-Euclidean geometry and relativity further unsettle an immediate identification of physical space with a prior mathematical geometry. Reichenbach's Philosophie der Raum-Zeit-Lehre (1928; English translation 1958) examines the relationship between geometry, coordinative definitions, and physical measurement. The philosophical consequence is that physical geometry involves an interpreted system of relations tested through measurement procedures. Neither a free-floating mathematical manifold nor an untheorized collection of observed distances suffices for physical spacetime ontology. The apparent simplicity of the classical container picture had obscured these layers of theoretical and empirical mediation.
From Positivism to Scientific Ontology
The question whether scientific theories describe reality or merely organize observations became increasingly explicit with nineteenth- and twentieth-century methodological debates. Duhem's La théorie physique: son objet et sa structure (1906; English translation 1954) presents physical theory as an ordered mathematical system that represents experimental laws. His discussion of testing shows why a failed prediction ordinarily challenges a network of assumptions rather than a single isolated proposition. The resulting holistic pressure complicates any direct argument from one experiment to the existence of one theoretical constituent.
Logical empiricist approaches sought to clarify the relation of theoretical language to observational confirmation. These programs differed among themselves, and their historical development cannot be captured by a simple assertion that twentieth-century philosophers denied unobservable entities. The lasting issue was how a theory acquires empirical meaning and what accepting its apparatus commits one to believing. In a mature scientific context, a commitment to the reliability of a measurement procedure can be justified differently from a commitment to the existence of every theoretical variable used to calculate the reading.
The later revival of scientific realism and naturalized metaphysics reoriented the debate toward the explanations supplied by successful theories and their continuity under scientific change. Its central question became whether success would be intelligible if theoretical descriptions never latched onto unobservable structures of the world. That question motivated competing answers: realist interpretations of approximate truth; pragmatic or empiricist accounts of predictive success; selective commitment to manipulable entities; and more demanding structural descriptions. Modern scientific ontology emerged as a field in which semantics, epistemology, experimental history, and physics are jointly necessary.
Quine and the Logic of Ontological Commitment
Quine's "On What There Is" (1948) gave analytic ontology an influential logical discipline. On a suitable regimentation of a theory into quantified logic, the objects over which bound variables must range for the theory to be true express its ontological commitments. A formulation such as "there exists an x such that x is an electron" commits that formulation, under its intended interpretation, to electron-like referents. The criterion is valuable because it asks philosophers to state the consequence of a theory's propositions instead of relying on unexamined grammatical appearances (Quine, 1948, pp. 32–33).
Yet the method does not settle which scientific formulation deserves literal interpretation. An equivalent empirical account may employ different mathematical objects or introduce a convenient representational parameter. The crucial question becomes whether an admissible paraphrase eliminates a commitment without losing explanatory and empirical content. The conceptual difficulty is that the standards of admissibility are substantive: preserving observational predictions is weaker than preserving counterfactual explanation, causal interpretation, and the structure of accepted scientific reasoning. Therefore the quantifier test is a test of commitments within a regimented theory, not an automatic test of physical existence.
The distinction is especially important in modern physics. A functional integral can quantify over field configurations, but it would be hasty to infer that every integration variable identifies an existing physical system. Coordinate labels may be necessary to specify a description while their numerical values have no invariant physical significance. Conversely, a physical claim may involve a real state or process even when a convenient formalism does not present it as a discrete object. Quine's contribution is a demand for interpretative clarity. Its productive extension requires a separate inquiry into how empirical evidence and explanatory practice warrant the semantic commitments that logical regimentation makes visible.
Scientific Realism and the Existence of Unobservables
Scientific realism maintains, in a broad family of versions, that successful science yields knowledge of some observable and unobservable aspects of a mind-independent world. Its ontological dimension concerns what there is, its semantic dimension concerns whether claims about theoretical entities are truth-apt, and its epistemic dimension concerns the warrant for belief. Separating these dimensions reveals how two philosophers may endorse the same experimental result while disagreeing about the truth of the theoretical story used to explain it. The distinction is central to contemporary analyses of realism (see Chakravartty, 2017).
The no-miracles argument begins with the impressive predictive and interventionist success of mature scientific theories. If their central descriptions bore no relevant relation to the real constitution of their target systems, their continuing success across varied conditions would appear surprising. The realist proposes that approximate truth, reference, or reliable representation of structure supplies the best explanation of this success. This is an abductive argument rather than a deduction. It depends on how alternative explanations are compared and on which parts of a complicated theory are responsible for reliable results.
One objection is that successful prediction can arise from idealizations, fitting practices, and systematic relations that do not correspond to a literal inventory of objects. A further objection appeals to historical theories whose central mechanisms are no longer accepted. Realist responses often become selective: warrant attaches to particular explanatory or experimental commitments rather than to every sentence of the currently favored theory. Such selectivity is rational only if its criteria can be specified prospectively, independently of knowing which theoretical claims later survived. Scientific ontology therefore inherits from realism a demand to explain success, and from its critics a demand to discriminate warranted entities from merely successful representational instruments.
Constructive Empiricism and the Limits of Belief
Bas van Fraassen's The Scientific Image (1980) presents constructive empiricism as an alternative to scientific realism. A theory may be interpreted literally, including its statements about unobservable objects, while acceptance of the theory requires belief only in its empirical adequacy. On this view, science aims to construct theories whose empirical substructures represent observable phenomena correctly. The stronger belief that the theory's account of unobservable nature is true is a distinct philosophical decision. Constructive empiricism thus differs from an instrumentalism that treats unobservable statements as meaningless or mere shorthand.
The observable–unobservable distinction generates difficulties. Whether something is observable depends on perceptual and physical capacities; microscopes, tracks, and indirect detection change what investigators can establish without making the philosophical category transparent. A distant planet and a subatomic interaction differ in their accessibility, but physical practice often connects direct and instrument-assisted detection through coherent chains of calibration. An empiricist may accept robust instrumental outcomes while withholding the further conclusion that the favored microscopic explanation is uniquely true. Realists object that some such withholding is epistemically artificial when multiple independent interventions converge on the same entity.
Constructive empiricism clarifies a decisive distinction between using a theory and believing its whole ontological interpretation. The position faces the task of articulating observation and empirical adequacy across sophisticated instruments and theories, but it exposes a genuine gap between successful application and an unrestricted metaphysical assertion. A defensible scientific ontology must argue across that gap. Whether the gap is closed for a particular class of entities depends on evidence stronger than the fact that a predictive formalism contains expressions naming them.
Intervention, Detection, and Entity Realism
Entity realism assigns particular importance to experimental control of putative entities. In Representing and Intervening (1983), Hacking argues that scientific experiments can justify belief in certain entities even where competing theories about their complete nature remain unsettled. The capacity to deploy electrons or other unobservables as stable causal instruments supports a form of realism grounded in practice. Cartwright's How the Laws of Physics Lie (1983) likewise challenges the literal truth of broad idealized law statements while defending causally significant features of particular experimental situations. These works shift attention from passive theoretical confirmation to active, repeatable transformation.
An intervention warrants ontological belief through an explanatory relation. The apparatus is designed using established properties; the intervention produces discriminating effects; independent methods can reproduce or constrain the interpretation; and competing explanations must account for the same detailed results. Such success provides more than a single observation of an anomalous trace. The J. J. Thomson electron case illustrates the importance of accumulating independent kinds of evidence rather than selecting one philosophically dramatic demonstration. Hacking's well-known emphasis on manipulation identifies a powerful source of confidence but does not entail that every successfully manipulated theoretical posit is fundamental.
A limitation remains: intervention is ordinarily described using background theories, and apparatuses contain their own idealizations. Moreover, some credible scientific entities cannot be deliberately manipulated, including many cosmological systems and unique historical processes. An intervention-only criterion would therefore exclude more than physics itself excludes. The appropriate conclusion is that reliable causal control is a strong, defeasible route to entity warrant, to be supplemented by independent detection, robust inference, and explanations that connect different regimes of measurement.
Historical Theory Change and Referential Continuity
History provides the most serious test of unqualified realist inference. Scientific representations have repeatedly displaced earlier accounts while retaining some empirical achievements. Laudan's "A Confutation of Convergent Realism" (1981) challenges the assumption that empirical success reliably tracks theoretical reference or approximate truth. Once a succession of once-successful theories is considered, the mere fact of current success supplies an incomplete reason to accept a theory's entire ontology. The pessimistic induction generalizes this historical instability to present theories, suggesting that currently favored categories too may be abandoned.
Realists respond by identifying what is preserved. Some invoke continuity in reference: an evolving description of electrons may retain its reference even as its theory of underlying properties changes. Others select experimentally established causal capacities or stable mathematical relations. Structural realists especially emphasize the persistence of relations through major conceptual transformations. These are promising responses, but selective continuity must be earned. If the favored enduring components can be identified only retrospectively, the defense risks becoming a rule that cannot guide current ontology.
Three questions clarify the dispute. Did a past theory genuinely refer to the same entity as a present theory, or did it share only some predictions? Did its supposedly successful component contribute indispensably to the success, or merely accompany other adequate machinery? Could contemporaneous evidence have singled out the retained component without hindsight? An answer can vary from case to case. A theory's partial failure need not defeat every belief grounded in it, while preservation of an equation need not establish preservation of an earlier material mechanism. Historical reconstruction therefore supports local, comparative assessments of physical entities rather than a uniform judgment about entire scientific eras.
Structural Realism and the Identity of Objects
Structural realism proposes that scientific knowledge may attach more securely to relations and organization than to the intrinsic natures of particular unobservables. Worrall's epistemic version recommends restraint concerning the natures of entities while preserving the structurally continuous parts of theories. Ontic structural realism, developed in different forms by Ladyman and other philosophers of science, makes a stronger metaphysical claim: relational structure may be ontologically prior to, or more fundamental than, conventionally individuated objects. Ladyman and Ross's Every Thing Must Go (2007) defends a naturalistic metaphysics responsive to contemporary physics and challenges traditional assumptions about independently self-subsistent microscopic individuals.
These versions answer different questions. Epistemic structural realism concerns what scientific success warrants us to know. Ontic structural realism concerns what the physical world fundamentally contains. Conflating them would infer a thesis about being from a limitation in knowledge. One might know only relational properties while objects with additional unknown characteristics still exist. Conversely, one might maintain the metaphysical priority of relations without claiming that a particular scientific equation reveals their final form.
Structural realism also faces a determination problem. A formal structure requires specification of a domain, relations, mappings, and physically significant constraints. If every relation can count as structural knowledge, the doctrine becomes empty. If the relations are tightly constrained by observation, symmetry, and counterfactual behavior, structural content can be explanatory. A stronger version of the view must explain how relata are individuated or why their individuality can be derived from the relational scheme. The modern quantum debate about permutation symmetry and non-individuality makes this an empirical as well as a conceptual pressure on an otherwise attractive philosophical position.
Models, Idealizations, and Representational Artifacts
Scientific reasoning commonly proceeds through models that isolate selected relationships. An ideal gas has mathematical properties that no perfectly actual gas possesses. A frictionless surface eliminates mechanisms present in real systems. A point charge simplifies electromagnetic calculations although real charge carriers must be described with greater care. A model can be extraordinarily accurate for its intended purposes because the idealization suppresses irrelevant variation. The success of the model warrants confidence in the features whose preservation accounts for the success, not automatically in the literal existence of every object assumed by its equations.
Cartwright (1983) foregrounds the gap between abstract laws and the complicated arrangements to which they are applied. Subsequent philosophy of scientific representation has analyzed similarity, structural mappings, inferential use, and interpretation as possible accounts of how models target phenomena. The ontological question is not exhausted by asking whether an idealized model resembles its target. One must specify the respects of comparison, the mapping from variables to quantities, the approximations involved, and the experiments that discriminate acceptable representations from poor ones.
This yields a useful rule: existence claims require an interpretation that survives removal of demonstrably representational conveniences. Suppose two models produce identical target predictions but assign different intermediate objects. The shared predictive structure may be well warranted, whereas the ontological status of the competing intermediaries remains unsettled. The rule does not mean that only observables are real; unobserved causes may be necessary to explain robust connections across different phenomena. It means that formal presence and physical commitment are distinct relations. A serious scientific ontology reconstructs the pathways between them case by case.
Particles, Identity, and Quantum Indistinguishability
Classical particles are commonly modeled as individuals with trajectories and intrinsic states. Quantum theory complicates this intuitive picture. In systems of identical particles, exchanging labels need not produce a distinct physical state. Bose–Einstein and Fermi–Dirac statistics reflect symmetrization requirements with no straightforward counterpart in the classical practice of assigning persistent individual labels. The resulting issue is metaphysical: do quantum particles lack individuality, possess a nonclassical individuality, or remain objects whose identity is described differently from the classical case?
Saunders (2006) investigates whether quantum particles count as objects and argues against a simple equation of permutation invariance with the abandonment of objecthood. The lesson extends beyond particles. An account of objects must specify which properties individuate them and whether those properties are invariant under physically appropriate transformations. Mere distinguishability by an arbitrary index is insufficient. If two formal descriptions differ only by a permutation that changes no physical situation, the index cannot carry the burden of establishing two differently identified physical worlds.
Quantum statistics is a reason to refine metaphysical categories, but it does not by itself decide between ontic structural realism, non-individualism, or alternative accounts of identity. Different interpretations place different weight on particle trajectories, state vectors, and observables. Moreover, ordinary detector events can be described as particle detections without implying that a fundamental ontology must consist entirely of classical miniature bodies. Ontological analysis must trace the inferential move from localized detection event to entity with stable identity and then ask what, if anything, licenses the stronger commitment.
Wave Functions, Measurement, and Alternative Ontologies
The quantum state assigns amplitudes from which experimental probabilities are calculated. Whether the wave function describes an objective physical field, an informational representation, a nomological structure, or some other feature is a central ontological controversy. Unlike a classical field over ordinary three-dimensional space, a many-particle wave function is standardly expressed over configuration space. A literal field interpretation therefore raises questions about the dimensionality and relation of its domain to familiar physical space. Calling the state "real" is only the beginning of an account of what kind of reality it represents.
The measurement problem sharpens the stakes. Linear quantum evolution can generate superpositions correlating a measured system with an apparatus, while experiments yield determinate recorded outcomes. Different proposed resolutions distribute ontology differently. Bohm's 1952 hidden-variable formulation gives a definite configuration and a dynamics guided by the quantum state. Collapse approaches alter the dynamics or provide additional rules for definite outcomes. Everettian approaches, developed in contemporary detail by Wallace (2012), interpret the universal quantum state as supporting emergent decohered branching structure. These proposals are not interchangeable metaphysical glosses: their explanatory resources and ontological commitments differ.
The philosophical standard is to identify the interpretation's primitives and explain why its dynamics produces the observed statistics and experienced records. It is also necessary to distinguish the mathematical possibility of an interpretation from decisive empirical support for it. The conventional quantum formalism generates extraordinarily successful predictions, but this fact does not adjudicate every interpretation of measurement. Quantum ontology therefore illustrates the broader thesis of this study: empirical success can strongly confirm structural and probabilistic claims while leaving multiple coherent pictures of the physical constituents of the world under investigation.
Bell, Nonseparability, and Physical Locality
Einstein, Podolsky, and Rosen (1935) challenged the completeness of quantum-mechanical description by analyzing correlations between previously interacting and spatially separated systems. Their argument depends on a proposed sufficiency criterion for elements of reality and on assumptions about the independence of distant physical systems. Bohr's response (1935) disputes the application of that criterion in quantum contexts and emphasizes the conditions under which physical quantities acquire experimental meaning. The exchange is a landmark because it turns the abstract question of completeness into an analysis of what can be inferred from correlations and separated measurements.
Bell's 1964 theorem reformulated a crucial part of the dispute. Under explicitly stated premises, certain local hidden-variable models satisfy constraints that quantum predictions can violate. Subsequent experiments have provided powerful evidence against broad classes of locally causal models satisfying the relevant statistical assumptions. The conclusion does not simply license the slogan that information travels faster than light or that quantum theory uniquely establishes one metaphysical interpretation. Which assumption must be revised depends on the precise conceptual framework, including locality, measurement independence, and the treatment of outcomes.
For scientific ontology, Bell's work establishes that the relations among physical events impose substantive constraints on proposed inventories of entities. A collection of separately intrinsic local states may fail to capture the empirically warranted correlation structure. The analysis therefore presses accounts of separability, causal explanation, and spatial individuation. It also demonstrates how a metaphysical dispute can generate experimentally assessable consequences after careful formal reconstruction. The existence of a distinction between ontology and testable theory does not prevent an ontological proposal from becoming vulnerable to evidence once its assumptions have been stated with sufficient precision.
Quantum Fields and the Limits of Particle Talk
Quantum field theory provides the principal framework for high-energy particle physics, but its ontological interpretation is far from settled. Scattering calculations naturally speak of particles being created, annihilated, and detected. In other settings, quantum fields or algebras of local observables supply a more systematic description. The fact that both vocabularies are indispensable in particular calculations does not determine which one names the ultimate constituents of reality. A philosophical interpretation must attend to the formalism actually used, its physical domain, and the role of idealized asymptotic states.
Doreen Fraser's analysis of interacting quantum field theories (2008) argues that a particle concept developed for free-field cases cannot straightforwardly be transported to interacting theories. This is a methodological warning against deriving an ontological inventory from the most convenient textbook cases. The existence of inequivalent representations and complications of localization likewise makes it difficult to identify field operators with classical fields at each spacetime point. Even if some version of a field ontology is defended, the field must be characterized under the relevant quantum and relativistic constraints rather than by analogy with a continuous classical substance.
The proper response is neither a wholesale rejection of particles nor a declaration that all talk of fields is merely instrumental. Localized detector excitations and asymptotic scattering states support a legitimate particle-oriented language within specified regimes. Correlation functions, field algebras, and invariant quantities may be better candidates for a more general framework. The philosophical challenge is to determine which facts are robust across these representations and where the interpretation involves genuinely different assumptions. Contemporary foundational reviews of quantum field theory emphasize that this determination remains an active field of inquiry.
Gauge Symmetry and Surplus Mathematical Structure
Many physical theories contain descriptions that differ mathematically while representing the same physical situation. Electromagnetic potentials provide a familiar example: appropriate gauge transformations change potential variables without altering gauge-invariant electromagnetic quantities. The ontology inferred from such a theory should not proliferate physically distinct worlds solely because redundant formal descriptions can be written down. A gauge orbit, rather than a single arbitrary representative, may be the relevant object of physical interpretation, although the details depend on the theory and on global or boundary conditions.
Gauge structure raises an important philosophical distinction. A symmetry may be interpreted as a redundancy of description, as a transformation relating genuinely distinct possibilities, or as an element whose status depends on how a theory's boundary conditions and reference systems are defined. It would be too crude to infer from the word "symmetry" alone that every transformed state is ontologically identical. The interpretation demands scrutiny of empirical observables, conserved quantities, allowable transformations, and global structure. In gauge theories with nontrivial topology, local formal equivalence can coexist with physically significant global distinctions.
This situation exemplifies the limits of syntactic commitment. Counting variables in a Lagrangian cannot by itself determine how many independent physical things the theory describes. Only after constraints, redundancy, and observables have been identified can variables be interpreted as distinct physical degrees of freedom. A sound scientific ontology therefore treats representational invariance as a substantive clue. Its task is not simply to discard formal structure, but to explain which aspects of the structure correspond to objective differences in possible physical arrangements.
Spacetime, General Relativity, and Background Structure
General relativity describes gravitation in terms of dynamical spacetime geometry, connecting curvature to distributions of matter and energy. This radically alters the Newtonian picture in which material motions are represented against a fixed geometric stage. The ontological question becomes whether spacetime points, metric fields, relations between events, or some structurally constrained combination should count as the basic physically real ingredients. The mathematical manifold is an indispensable part of the formal description; its bare points, however, are not automatically individuated as distinct physical locations independently of the fields and relations instantiated on it.
The hole argument illustrates the issue. Diffeomorphism-related models can be understood as representing one physical situation rather than distinct possibilities differing only in which bare manifold points receive which field properties. The argument relies on substantive interpretive assumptions, and it has generated competing accounts of sophisticated substantivalism and relationism. Its enduring contribution is to show that a theory's coordinate representations and the ontology of spacetime must be distinguished through a principled account of physical equivalence.
Physical measurements of interval, curvature, gravitational lensing, and gravitational waves support the dynamical efficacy of general-relativistic structure. They do not force an unambiguous answer to every philosophical question about spacetime's most fundamental metaphysical category. The lesson parallels gauge theory: the observable and explanatory content of a geometrical field theory is realized through invariant and relational properties, while competing ontologies interpret the bearers of those properties differently. A mature account must preserve the precise content of physical tests without silently converting an underdetermined metaphysical thesis into an experimental fact.
Systems, Composition, and Emergent Physical Kinds
Physical systems are frequently more than arbitrary sets of primitive constituents. A crystal, fluid vortex, plasma, or superconducting material exhibits regularities that depend on collective organization and boundary conditions. Scientific ontologies must accommodate these systems even when the most basic dynamical theory is expressed in terms of other variables. Their existence as physically organized individuals can be supported by repeatable interventions, stable behavior, and cross-context identification without requiring that each be fundamental in the sense of admitting no deeper explanation.
Questions of composition become especially important when the system's explanatory properties are not transparently derivable by simply listing those of isolated parts. Statistical mechanics and condensed matter theory use collective variables, phase transitions, and idealized limits to characterize such behavior. Batterman's account of asymptotic reasoning explains why the elimination of microscopic detail can be essential to understanding certain universal patterns. The explanatory legitimacy of this method does not imply that organized systems float free of their material constitution; it shows that successful physical explanation is sensitive to levels and limits.
The ontological question is whether a higher-level entity has conditions of persistence, causal organization, and law-governed behavior robust enough to warrant recognition. Multiple realizability and universality suggest that higher-level kinds may be stable across materially different configurations. Reductionists can recognize this stability while arguing that all higher-level facts depend on fundamental physical facts. Emergentists may add stronger claims about novel causal or ontological organization. The monograph's framework requires these additional theses to be specified, since independent scientific usefulness, causal efficacy, and metaphysical fundamentality are distinct dimensions.
Effective Theories and Scale-Dependent Ontology
Effective theories characterize a domain at a chosen range of energies, lengths, or explanatory scales. Their parameters and degrees of freedom are tuned to the phenomena that matter within that regime. The theory may integrate out or suppress degrees of freedom that become relevant at shorter distances, and the resulting description may nevertheless support extremely precise predictions. The existence of an effective description is therefore compatible with the expectation of a more detailed underlying theory. The question is how its entities and structures should be interpreted while the more encompassing account remains incomplete.
An overly restrictive fundamentalism would count only entities in the final microscopic theory, a theory we do not presently possess. An indiscriminate realism would grant equal physical status to every term in every successful effective calculation. Both miss the graded nature of warranted claims. Some effective entities, such as collective excitations in condensed matter, support stable intervention and detection within a specified domain. Other quantities are bookkeeping devices introduced by approximation schemes. Their status cannot be read directly from the scale at which they enter the equations.
Renormalization and universality are especially instructive. Distinct microscopic arrangements can yield the same macroscopic scaling behavior, suggesting an objective robustness not tied to one microscopic mechanism. This supports realism about patterns and relations whose empirical content survives changes in detailed assumptions. It also limits overconfident claims that a successful effective theory has disclosed a final list of fundamental physical things. Scientific ontology becomes scale-sensitive when it distinguishes the level at which an entity is reliably identifiable from claims about what underlies that level.
Inference to Existence and Degrees of Ontological Warrant
The preceding arguments permit a systematic reconstruction of ontological inference. A proposal that some entity exists first requires a clear statement of the entity's identity conditions or relevant physical role. The representation must then connect those conditions to evidence: characteristic detector responses, discriminating predictions, independent measurements, stable causal intervention, or explanatory integration across phenomena. Alternative representations must be assessed, particularly those that preserve empirical adequacy without the proposed entity. Finally, the interpretation must survive relevant theoretical transformations, idealization analyses, and distinctions between local and fundamental description.
These requirements do not form a universal numerical score. Experimental interventions, theoretical unification, and model robustness can be mutually supportive or dependent on the same background assumptions. Independence is especially important: five observations arising from a common calibration error are not equivalent to five genuinely different detection mechanisms. A new entity gains warrant when its proposed causal and structural role explains a range of phenomena under conditions that expose the interpretation to possible failure. Evidence for a stable phenomenon may precede agreement about the best description of its ultimate bearer.
Three degrees of commitment illustrate the point. One may be strongly warranted in believing that a detector interaction occurred, reasonably warranted in identifying a particle-like excitation, and still unsettled about whether the relevant physical ontology is fundamentally particulate, field-theoretic, or structural. These degrees correspond to different propositions, so their coexistence involves no contradiction. This account also makes room for scientific progress: new experiments can strengthen a local claim, while theoretical advances can change the preferred global metaphysical interpretation without invalidating the former. Ontological discipline consists in matching the strength and target of belief to the evidence actually obtained.
Comparative Evaluation and Outstanding Objections
A global realism about entire theories has the virtue of taking scientific success seriously, but it encounters idealization and historical replacement. Constructive empiricism avoids overcommitment while inviting questions about why highly manipulable and independently detected unobservables should remain objects of suspended belief. Entity realism offers a compelling account of experimental warrant while requiring supplementation for inaccessible cosmological cases and for theory-mediated properties. Structural realism addresses continuity and quantum individuation, yet must provide a determinate concept of physical structure and a theory of how relations obtain. Naturalized metaphysics prevents philosophy from drifting away from established physics, while still needing a philosophical account of interpretation and the choice among empirically equivalent descriptions.
A central objection to the synthesis advanced here is that it may merely rename selectivity without identifying an ontological principle. Its answer is that selectivity becomes determinate once the target proposition, transformation invariance, evidential independence, and explanatory role are specified. The method yields conditional judgments rather than a universal inventory: if a hypothesized entity has stable empirically confirmed capacities that distinguish it from empirically equivalent rivals, commitment is stronger. When a theory's alternative formulations share all relevant empirical and explanatory relations, the additional distinctions between their posits may lack warrant. This is a substantive consequence, not a preference for philosophical caution.
A second objection is that inference to the best explanation already provides an adequate general solution. Yet what counts as an explanation depends on background ontology, standards of simplicity, and the role assigned to idealized mechanisms. Inference to the best explanation can be an indispensable form of reasoning without making the outcome mechanically unique. A third objection maintains that only fundamental physics should decide ontology. This overlooks the explanatory and experimental stability of nonfundamental systems and mistakes a conjecture about ultimate dependence for a completed theory of material existence. The comparative result favors disciplined pluralism about kinds of physical entity, under a common standard of explicit interpretation and accountable evidence.
Open Problems in Contemporary Scientific Ontology
Several unresolved questions bear directly on the ontology of physics. Quantum mechanics lacks a universally accepted resolution of the measurement problem. Competing interpretations make different claims concerning the status of wave functions, outcomes, trajectories, and branching structures. Quantum field theory complicates the localization and individuality of particles and challenges the assumption that field operators can be interpreted like classical fields. The relation between algebraic structures, scattering calculations, and physical ontology requires further work, particularly in interacting regimes. These issues concern the interpretation of well-supported formalisms as much as the search for new experimental results.
A second family of open problems concerns spacetime and unification. General relativity treats geometry dynamically, while quantum descriptions of matter commonly presuppose background structures adapted to different mathematical methods. An adequate theory of quantum gravity may change the ontological standing of spacetime, localization, and fundamental degrees of freedom. Existing proposals offer distinct possibilities, and the available evidence does not establish one final metaphysical picture. Similarly, the physical status of gauge-invariant observables, nonlocal relations, and global topological features demands a careful connection between formal theory and what can actually be measured.
A third problem concerns ontology across scales. Effective field theories, thermodynamic universality, and emergent phases support robust kinds without settling whether their apparent autonomy entails strong emergence. A fourth concerns the epistemology of discovery: how should experiments designed within one interpretation discriminate among interpretations that are observationally equivalent under currently accessible conditions? The most productive response is to make competing commitments explicit, identify possible discriminators, and preserve results that remain warranted across alternatives. Physical ontology remains a research program precisely because scientific success and metaphysical completion are different kinds of achievement.
Philosophical Synthesis
The historical and argumentative evidence supports an interpretation-centered account of scientific ontology. A physical ontology is an articulated set of existence and dependence claims whose justification depends on how theoretical structures refer to and explain physical phenomena. Its content cannot be extracted from ordinary language alone, since physical theories employ idealizations, unobservable posits, redundant variables, and mathematically equivalent descriptions. Nor can its content be determined by a single philosophical maxim about the primacy of objects, fields, or relations. Ontological conclusions require explicit representation, empirical testing, and a reasoned treatment of rival formulations.
The analysis yields five connected results. First, existence and fundamentality are distinct; reliable higher-level physical kinds need not be primitive. Second, a formal commitment and a warranted commitment are distinct; quantified variables acquire physical significance through interpretation. Third, the best evidence for unobservables combines independent detection, discriminating explanation, and in suitable cases experimental intervention. Fourth, invariance under changes of representation is a powerful diagnostic, though identifying the relevant transformations itself requires physical judgment. Fifth, historical continuity supports realism about successful components only when the components are identified by principled features rather than by retrospective survival alone.
The central philosophical consequence is that physics provides increasingly constrained descriptions of the world without guaranteeing that one current representation gives the final inventory of being. This conclusion recognizes the ontological authority of successful science while preserving philosophy's task of analyzing interpretation, identity, reference, and explanatory dependence. Particles, fields, systems, and structures can all enter a responsible ontology when the evidence and representation justify their particular status. The task of scientific ontology is to state that status precisely, not to force all physical phenomena into the grammar of one favored metaphysical category.
Conclusion
Physical entities are established through a conjunction of theoretical meaning and empirical relations. Their status depends on the content of a physically interpreted theory, the evidence supporting its explanatory and interventionist claims, and the stability of those claims across scientifically justified transformations. Aristotle's inquiry into natural change, Newtonian mechanics, field theory, relativity, and quantum physics each supply historically distinctive resources for addressing this problem. The history illustrates the strength of progressively refined physical knowledge and the changing categories through which that knowledge is represented.
Realism correctly insists that scientific success often warrants belief about unobservable nature. Empiricism requires that such belief be argued rather than assumed. Entity realism reveals the significance of experimental practice, while structural realism identifies the importance of relational continuity and invariance. Quantum theory, gauge structure, and effective field theories show why none of these resources alone suffices to identify every physical kind. The resulting synthesis establishes a method for assigning differentiated ontological warrant: identify the precise proposition, articulate its representational content, evaluate the causal and experimental evidence, and test its resilience under reasonable alternatives.
The subject therefore remains a foundational investigation of physical being. Its philosophical completion at any historical moment consists in establishing justified distinctions and explicit consequences, while new physical theories and experiments continue to alter the field of defensible commitments. Scientific ontology is most rigorous when its conclusions are proportionate to the content of the science it interprets.
References
- Aristotle. (c. fourth century BCE). Physics, Book II, part 1 (R. P. Hardie & R. K. Gaye, Trans.). Internet Classics Archive. Text of Book II.
- Batterman, R. W. (2001). The devil in the details: Asymptotic reasoning in explanation, reduction, and emergence. Oxford University Press. https://doi.org/10.1093/0195146476.001.0001.
- Bell, J. S. (1964). On the Einstein Podolsky Rosen paradox. Physics Physique Fizika, 1(3), 195–200. https://doi.org/10.1103/PhysicsPhysiqueFizika.1.195.
- Bohm, D. (1952a). A suggested interpretation of the quantum theory in terms of "hidden" variables. I. Physical Review, 85, 166–179. https://doi.org/10.1103/PhysRev.85.166.
- Bohm, D. (1952b). A suggested interpretation of the quantum theory in terms of "hidden" variables. II. Physical Review, 85, 180–193. https://doi.org/10.1103/PhysRev.85.180.
- Bohr, N. (1935). Can quantum-mechanical description of physical reality be considered complete? Physical Review, 48, 696–702. https://doi.org/10.1103/PhysRev.48.696.
- Bricker, P. (2014). Ontological commitment. In The Stanford Encyclopedia of Philosophy. Metaphysics Research Lab, Stanford University. https://plato.stanford.edu/entries/ontological-commitment/.
- Cartwright, N. (1983). How the laws of physics lie. Oxford University Press. https://doi.org/10.1093/0198247044.001.0001.
- Chakravartty, A. (2017). Scientific realism. In The Stanford Encyclopedia of Philosophy (revised June 12, 2017). Metaphysics Research Lab, Stanford University. https://plato.stanford.edu/entries/scientific-realism/.
- Duhem, P. (1954). The aim and structure of physical theory (P. P. Wiener, Trans.). Princeton University Press. (Original French work published 1906.) Bibliographic record.
- Einstein, A., Podolsky, B., & Rosen, N. (1935). Can quantum-mechanical description of physical reality be considered complete? Physical Review, 47, 777–780. https://doi.org/10.1103/PhysRev.47.777.
- Fraser, D. (2008). The fate of "particles" in quantum field theories with interactions. Studies in History and Philosophy of Modern Physics, 39(4), 841–859. https://doi.org/10.1016/j.shpsb.2008.05.003.
- Frigg, R., & Nguyen, J. (2026). Scientific representation. In The Stanford Encyclopedia of Philosophy (substantive revision May 20, 2026). Metaphysics Research Lab, Stanford University. https://plato.stanford.edu/entries/scientific-representation/.
- Hacking, I. (1983). Representing and intervening: Introductory topics in the philosophy of natural science. Cambridge University Press. https://doi.org/10.1017/CBO9780511814563.
- Kuhlmann, M. (2026). Quantum field theory. In The Stanford Encyclopedia of Philosophy (substantive revision September 9, 2026). Metaphysics Research Lab, Stanford University. https://plato.stanford.edu/entries/quantum-field-theory/.
- Ladyman, J. (1998). What is structural realism? Studies in History and Philosophy of Science Part A, 29(3), 409–424. https://doi.org/10.1016/S0039-3681(98)80129-580129-5).
- Ladyman, J. (2023). Structural realism. In The Stanford Encyclopedia of Philosophy (substantive revision May 18, 2023). Metaphysics Research Lab, Stanford University. https://plato.stanford.edu/entries/structural-realism/.
- Ladyman, J., & Ross, D. (with Spurrett, D., & Collier, J.). (2007). Every thing must go: Metaphysics naturalized. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199276196.001.0001.
- Laudan, L. (1981). A confutation of convergent realism. Philosophy of Science, 48(1), 19–49. https://doi.org/10.1086/288975.
- Newton, I. (1687). Philosophiae naturalis principia mathematica, Definitions and Scholium. London. Newton Project critical transcription.
- Quine, W. V. O. (1948). On what there is. The Review of Metaphysics, 2(1), 21–38. https://www.jstor.org/stable/20123117.
- Reichenbach, H. (1958). The philosophy of space and time (M. Reichenbach & J. Freund, Trans.; R. Carnap, Introduction). Dover Publications. (Original German work published 1928.) Bibliographic record.
- Saunders, S. (2006). Are quantum particles objects? Analysis, 66(1), 52–63. https://doi.org/10.1093/analys/66.1.52.
- van Fraassen, B. C. (1980). The scientific image. Clarendon Press/Oxford University Press. https://doi.org/10.1093/0198244274.001.0001.
- Wallace, D. (2012). The emergent multiverse: Quantum theory according to the Everett interpretation. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199546961.001.0001.
- Worrall, J. (1989). Structural realism: The best of both worlds? Dialectica, 43(1–2), 99–124. https://doi.org/10.1111/j.1746-8361.1989.tb00933.x.