Dispositions and Powers

Philosophical Overview: The Reality of What Can Happen

A pane of glass remains fragile while it sits intact in a display case. A magnet may attract an iron object even when no iron object lies nearby. A trained musician retains a capacity to play during an interval of silence. These examples exhibit a common philosophical structure: a claim about the present condition of a thing bears upon outcomes that need not occur at the time of the claim. The assertion that the glass is fragile is intelligible before, during, and after any particular impact. It concerns something possessed rather than merely an event observed. What gives such assertions their truth, and what gives the capacity its reality, constitute the central questions of the metaphysics of dispositions and powers.

The inquiry is foundational because a wide range of ordinary and scientific discourse relies on the distinction between possession and exercise. We describe water as capable of dissolving salt, a surface as electrically conductive, a drug as therapeutically active, a person as skilled, and a material as elastic. Some descriptions are probabilistic, some depend on elaborate environmental conditions, and some designate highly stable regularities. Yet they ordinarily license explanation and prediction beyond a simple enumeration of past occurrences. A power can help explain why a process happened, why a process was prevented, and why a process would have occurred under suitably different circumstances.

The philosophical difficulty begins when this explanatory language is asked to disclose its ontological commitments. The categoricalist can propose that a thing's actual qualitative or structural constitution, together with laws and circumstances, supplies the full explanation. The dispositional essentialist holds that at least certain properties are essentially directed toward characteristic manifestations, so that their modal profile belongs to what they are. A third approach understands a property as both qualitative and powerful, attempting to overcome the apparent opposition between an actual way of being and an ability to act (Martin & Heil, 1999; Heil, 2003; Bird, 2007). These views agree on many empirical descriptions while disputing the structure of the world that makes those descriptions true.

The guiding questions are therefore discriminable. There is a semantic problem about the meaning and truth conditions of a disposition ascription, a metaphysical problem about the entity or property that makes it true, a causal problem about how such an entity contributes to events, and a modal problem about its bearing on possibility and necessity. There is also an epistemic problem: experiments reveal manifestations in selected situations, yet the claims they support reach beyond those situations. Confusing these questions has sustained much of the dispute. A failed linguistic reduction does not automatically establish ontologically fundamental powers; a successful account of causal mechanisms does not automatically eliminate disposition predicates.

The argument developed here treats dispositions as directed, context-sensitive profiles of possible manifestation grounded in the actual constitution and relations of their bearers. This is initially a framework for analysis, not a declaration that the grounds are themselves irreducibly powerful. Whether grounds must ultimately have dispositional essences is a further thesis requiring independent argument. By separating what an ascription commits us to from competing theories of its grounds, we can identify genuine disagreements without mistaking shifts of vocabulary for metaphysical discoveries (Mumford, 1998; Molnar, 2003; Vetter, 2015).

Conceptual Architecture: Bearers, Bases, Stimuli, and Manifestations

The typical attribution of a disposition involves a bearer, such as a glass, a property ascribed to that bearer, such as fragility, and a manifestation, such as breaking. A stimulus condition, such as an impact of an appropriate magnitude, identifies a circumstance in which the manifestation may occur. The glass's microstructure is a possible causal basis of its fragility. The environment influences whether that basis produces its characteristic result. This elementary schema contains terms whose distinction is indispensable.

A manifestation is an event, process, or resulting state through which a disposition is exercised. A disposition may persist across an indefinitely large number of opportunities for manifestation without manifesting on any of them. The possession of a power and its exercise belong to different aspects of a thing's history. A scientist who infers a substance's solubility from its molecular constitution can ascribe the power while observing no dissolution. Conversely, witnessing dissolution does not settle whether the property should be individuated by that manifestation alone, or by that manifestation together with particular triggering circumstances.

The causal basis must be distinguished from the disposition itself. One view identifies the disposition with the basis: fragility is a structural feature specified at an appropriate level. Another holds that the disposition depends upon, but is not identical with, that basis. Still another regards a power as fundamental and so resists the demand for a categorical basis of a distinct kind (Prior, Pargetter, & Jackson, 1982; Molnar, 1999; Mumford, 1998). The relationship between a property's nature and its bearer is also contested. An intrinsic disposition depends entirely on the bearer as considered in itself; an extrinsic one depends on the wider arrangement. A key's ability to open a particular lock, an object’s legal authority, and the vulnerability of one system to another invite careful separation of relational capacities from the intrinsic capacities participating in them (McKitrick, 2003).

Manifestations often admit degrees and multiple realizations. Fragility may be higher or lower, a solution may dissolve quickly or slowly, and a skilled performer may succeed across a wider or narrower range of conditions. A binary formula of the form “if stimulated, then manifest” compresses graded profiles into an on/off category. Such compression can be useful in controlled cases but should not dictate the metaphysics of every disposition. Manley and Wasserman's work on gradable disposition ascriptions and Vetter's account of potentiality highlight the need to explain degrees, context, and the range of possible manifestations, rather than assuming that every disposition has one privileged triggering event (Manley & Wasserman, 2008; Vetter, 2015).

A capacity may also be joint. The power of two chemicals to react is exercised in their interaction, while the ability of a team to coordinate may depend on a structure distributed among individuals. This does not imply that no participating individual has powers. It means that the complete manifestation condition depends upon an organized plurality. Vetter (2015, ch. 4) develops the distinction between joint, extrinsic, and iterated potentialities, which clarifies why ascriptions to isolated bearers can be contextually incomplete.

A workable conceptual reconstruction therefore distinguishes six questions: Who or what bears the disposition? What present constitution or relation grounds it? Toward what outcome is it directed? Through which conditions is it activated or exercised? Which additional factors enable or prevent the outcome? What modal claim does the ascription warrant? These questions can be answered at multiple descriptive levels without committing the investigator to six different ontological entities. They define the explanatory tasks any successful theory must perform.

Aristotle: Dunamis, Energeia, and the Structure of Change

Aristotle provides the decisive classical point of departure. In Metaphysics Θ, he investigates δύναμις (dunamis) in relation to ἐνέργεια (energeia) and ἐντελέχεια (entelecheia), commonly rendered through the families of “capacity,” “potentiality,” “actuality,” and “fulfillment.” Translation matters. Dunamis in Θ.1–5 frequently designates a specific capacity to initiate or undergo change, whereas the broader potentiality–actuality discussion of Θ.6–10 concerns being and actuality beyond motion alone. Rendering the entire book with one English term obscures this progression (Aristotle, Metaphysics Θ.1, 1045b27–1046a19; Θ.6, 1048a25–1048b9).

The narrower definition identifies an active capacity as a principle of change in another thing or in its possessor insofar as it is other. The formulation marks a distinction between active and passive power. An agent can have the capacity to heat; an object can have the capacity to be heated. Their corresponding activities are intelligible through a relation between what acts and what can undergo the relevant process. Aristotle further distinguishes rational powers, which can be exercised toward contraries, from nonrational powers directed characteristically toward a determinate outcome. Medical knowledge can be used with respect to health and its contrary; the heating power of something hot is treated differently (Metaphysics Θ.1–2, 1046a10–1046b24).

This is an account of organized activity rather than a catalog of dormant possibilities. Aristotle's rational capacities require an additional determinant, such as choice or desire, to explain why one of several possible outcomes occurs. For nonrational capacities, the appropriate proximity of agent and patient plays a more direct role (Metaphysics Θ.5, 1047b31–1048a24). The distinction anticipates later problems of triggering and manifestation, but its embedding in an account of substance, form, and actuality makes it historically distinct from contemporary property metaphysics.

The Megarian challenge, which Aristotle confronts in Θ.3, identifies having a power with exercising it: a person can build only when building. Aristotle objects that this collapses the difference between a skill's possession and its exercise and produces implausible consequences concerning changes in ability and the existence of unactualized capacities (Metaphysics Θ.3, 1046b29–1047a24). The point has enduring significance. To recognize an ability as genuinely possessed is to allow its existence through intervals in which it remains unexercised. A historical continuity of the bearer and its relevant organization can support a real attribution even in the absence of a contemporaneous manifestation.

Aristotle also resists reducing every potentiality to the mere logical possibility of an event. A block can be potentially a statue only under material and practical qualifications; the potential is informed by what the thing is and by the conditions in which transformation can occur. The relation between potentiality and actuality involves asymmetry: actuality is prior in account and, in important senses, prior in substance and completeness (Metaphysics Θ.8, 1049b4–1051a3). Modern theories that make fundamental powers ontologically primary cannot simply invoke Aristotle as a precursor with identical commitments. They inherit a problem of change and directed possibility while revising his account of substance and actuality.

In Physics III.1, the analysis of motion as the actuality of the potential insofar as it is potential adds another distinction: the unfolding of a change cannot be equated with an already completed state. A glass's actual breaking, the process of breaking, and the prior capacity to break occupy different explanatory positions (Aristotle, Physics III.1, 201a9–201b15). The modern language of manifestation gains precision when it retains this difference between possession, activity, and outcome.

Early Modern Transformations: Power, Essence, and Experience

The early modern history of powers is a history of transformations rather than a linear advance from “occult qualities” to modern science. Rival systems gave power different places in the explanation of bodies, minds, and causation. Descartes' mechanical approach privileged extension, motion, and lawful interaction in physical explanation, while seventeenth-century debates about active force challenged the thought that extension and local motion exhaust corporeal reality. The question was whether matter can act through its own intrinsic principles and how such action is related to divine conservation and laws.

Locke's An Essay Concerning Human Understanding provides an influential analysis of the idea of power. In Book II, chapter XXI, the mind derives the idea from observing changes in things and recognizing possibilities of producing or receiving change. Locke distinguishes active and passive powers and regards many of the properties by which substances are known as powers to produce ideas or effects. His discussion of primary and secondary qualities also warns that the manifest sensory character of an effect cannot be taken as a transparent disclosure of the causal organization of its source (Locke, 1690, II.viii; II.xxi). A dispositional description is thus entangled with epistemology: inquiry may identify what a substance can do without penetrating the real essence that grounds the capacity.

Leibniz's analysis of active force gives a different answer. In New Essays on Human Understanding and the associated metaphysical writings, activity is central to the constitution of substance. The distinction between active and passive faculties is not exhausted by a mere inventory of observed transitions, while the account of force is embedded in a theory of simple substances and their expression. Leibniz's dynamics also contrasts force with the geometrical description of motion. His system cannot be assimilated directly to twentieth-century causal-power realism, but its thesis that explanatory activity belongs to the nature of things remains a major alternative to an ontology whose fundamental properties are wholly inert (Leibniz, 1765/1996, II.xxi; see also the Monadology, §§11–18).

Spinoza offers a further configuration. In Ethics III, propositions 6–8, a thing's striving (conatus) to persevere in being is presented as its actual essence insofar as it expresses the power by which it exists and acts. The relation between essence and activity is constitutive within Spinoza's necessitarian metaphysics. Yet his doctrine of the modes and the necessity of the divine nature diverges sharply from contemporary proposals that treat particular powers as independent causal ingredients potentially combining in open-ended ways (Spinoza, 1677/2002, IIIp6–p8). Comparative study is illuminating precisely because these authors give different answers to what makes an active nature intelligible.

The philosopher of science must also distinguish seventeenth- and eighteenth-century “powers” from a claim that every scientifically respectable explanation names a primitive productive force. A science can investigate relations among forces, interactions, and structures without solving the metaphysical issue of whether the properties mentioned possess their causal roles essentially. The historical vocabulary of force, energy, capacity, propensity, and power shifts across disciplines. Conceptual reconstruction must track those shifts before drawing ontological conclusions.

Hume, Necessity, and the Challenge to Productive Power

David Hume reorganizes the debate by asking what idea we possess when we attribute necessary connection or efficacy to causes. In A Treatise of Human Nature I.iii.14, he groups terms such as efficacy, power, force, and productive quality, then demands an experiential source for the idea of necessity. The experience of a single event's succession by another displays contiguity and succession, but no perceptible necessary tie. Repetition generates a customary transition in thought, and Hume explains the felt necessity through the mind's acquired expectation (Hume, 1739–1740, T 1.3.14.4–24).

Hume's challenge has two layers often run together. The epistemological question is whether observation provides access to a necessary tie in objects. The semantic and metaphysical question concerns what a claim of necessary connection could mean and whether an objective productive connection exists. A realist about powers may concede that the underlying connection is not directly perceived while arguing that it is warranted by inference to the best explanation, theoretical integration, or scientific intervention. Such a response answers a different question from the claim that an observer can literally see metaphysical necessity.

In the Enquiry Concerning Human Understanding, section VII, Hume returns to the origin of the idea of necessary connection and to the apparent inability to discover power simply from the sensible qualities of an object. His skeptical analysis challenges claims to infer an effect a priori from the bare conception of its cause. A dispositional essentialist replies that genuine property essences constrain possible behavior, but must explain how those essences are identified and why this explanation offers knowledge beyond contingent empirical regularities. It is insufficient to relabel the regularity “power” and declare necessity thereby explained.

The Humean framework has resources of its own. A mosaic of particular facts, on certain later reconstructions, supports lawlike generalization without primitive necessary connections. Disposition predicates may then be true because of actual categorical bases and the laws, regularities, or counterfactual patterns that obtain in a world. Powers metaphysics contends that such an arrangement leaves unexplained why the properties contribute as they do. That contention is substantive only if “explanation” is specified: an account can demand causal mechanisms, nomological compression, grounding, or metaphysical necessitation, and success at one task does not guarantee success at all.

The enduring result of the Humean challenge is methodological. One cannot move uncritically from successful prediction or the feeling of necessity to a demonstrated metaphysical production relation. Neither can the limits of direct observation by themselves demonstrate that scientific dispositional attributions are unreal. What remains is a comparative evaluation of theoretical commitments and explanatory results. The later controversy over conditional analyses can be understood as one attempt to keep the truth of a disposition claim while avoiding ontologically primitive powers.

Dispositions in Twentieth-Century Philosophy: The Logical Turn

The modern dispute was shaped by a linguistic and analytical ambition: to explain the role of disposition predicates without postulating hidden occurrences that resemble the events they help explain. Gilbert Ryle's The Concept of Mind (1949), especially “Knowing How and Knowing That” and “Dispositions and Occurrences,” treats many mental predicates as dispositional. To say that someone knows a language, is intelligent, or is irritable need not report a private episode continuously occurring behind behavior. Such a statement places what the person does, could do, or characteristically does in an inferential network of conditions, activities, and performances. Ryle's treatment belongs within a broader critique of the attempt to understand every mental term as the name of a private inner event.

The insight is that the grammar of a capacity differs from the grammar of its exercise. The statement that a student can solve quadratic equations remains true while the student is asleep. The statement that the student is solving one is episodic. Their difference is neither a peculiarity of human minds nor proof that abilities are necessarily immaterial. It concerns the logical role of attributions that survive nonmanifestation. This anticipates a broad theory of dispositions while remaining distinct from a realist account of fundamental natural properties (Ryle, 1949).

A standard conditional approach represents a disposition schematically: an object x is disposed to manifest M in condition C if and only if x would manifest M were C to obtain. For a fragile glass the imagined stimulus is an appropriate impact and the manifestation is breaking. In a possible-worlds reading the claim involves a counterfactual relation: in sufficiently relevant nearby worlds where the stimulus occurs, the response occurs. This formulation is attractive because it expresses the forward-looking content of an attribution in terms of behavior under hypothetical circumstances.

Its virtues are analytical economy and close connection with observation. If a disposition could never lead to a distinctive result under any pertinent test, what content would remain in calling the object disposed? The conditional can also explain ordinary inference from possession to expected manifestations when conditions are controlled. Yet its apparent simplicity conceals significant assumptions: that the triggering condition is specifiable independently of the disposition, that the disposition persists through the test, that no factors interfere with its exercise, that an actual manifestation uniquely reveals the disposition, and that the modal similarity relation correctly tracks the phenomenon.

The philosophical literature already recognized complications in the relationship between a disposition and its causal basis. Prior, Pargetter, and Jackson (1982) investigate whether a disposition must have a causal basis and how those bases relate to dispositional properties. Mellor (1974) argues in defense of dispositions as a legitimate explanatory category. Mumford (1998) develops a sustained ontology of dispositions while challenging reductions that would dissolve them into distinct categorical grounds. The twentieth-century debate thus moves from logical analysis to questions about what in the world explains the truth of the relevant modal claims.

The distinction between an analytic reduction and a metaphysical reduction is decisive. A philosopher may hold that fragility is entirely grounded in a microstructure even if the meaning of “fragile” cannot be captured by a finite list of material descriptions. Equally, failure to define fragility by one counterfactual does not imply that it is metaphysically fundamental. Arguments must specify which thesis the purported counterexample defeats.

Finks: When Testing Changes the Power

C. B. Martin's “Dispositions and Conditionals” (1994) presents the classic fink objection. Imagine a live wire whose condition would be changed instantly by a mechanism whenever someone attempted the relevant test. Prior to contact the wire possesses the electrical disposition in question. The attempt to elicit its manifestation removes that disposition, and so the expected response never appears. Conversely, a device might arrange for an otherwise nonlive wire to become live precisely when the test begins. A simple counterfactual tying the present possession of a disposition to what would occur following a stimulus yields the wrong attribution in either direction.

The central difficulty concerns temporal dependence. Possession is evaluated at a time, while the counterfactual test introduces a temporally extended sequence. A condition may hold at the moment of evaluation and fail during the counterfactual sequence because the sequence itself alters the bearer. The formula “D(x,t) if and only if C would bring about M” tacitly assumes an invariance that cannot be guaranteed by the meaning of D. Changing properties do not violate ordinary metaphysical principles merely because a proposed analysis cannot accommodate their change (Martin, 1994).

David Lewis (1997) accepts the problem and attempts to improve the conditional analysis. His proposal identifies a suitable causal basis of a disposition and includes a condition that the basis persist for an adequate interval after stimulation. The manifesting process is then evaluated under preservation of its enabling ground. This is more discriminating than the simple conditional because a response prevented by instantaneous loss of its ground does not falsify the original attribution. It also illuminates a genuine relation between dispositions and causal mechanisms: the counterfactual must hold fixed something relevant to the capacity.

Lewis's repair creates fresh questions. First, not every disposition has a neatly identifiable basis whose persistence is uncontroversial. Second, the distinction between a basis and the wider enabling environment matters. Third, keeping a basis fixed may be inconsistent with the nature of some manifestations, since the exercise of a power can alter its bearer. A fuel's capacity to burn is exercised through transformations that destroy the precombustion arrangement. One therefore needs a principled account of how much of the basis remains fixed, for how long, and relative to which parts of the manifestation process.

There is also a counterfactual-selection issue. The conditions under which a disposition would manifest must be sufficiently specific to exclude interventions that disable it; but if these conditions are identified by saying that the manifestation will not be prevented, the analysis threatens circularity. A disposition ascription may remain useful precisely when no independent, exhaustive description of “favorable conditions” is available.

The fink problem establishes a fundamental point about capacities: tests are causal interventions in the world, and an intervention can change the property it is meant to reveal. Any semantics of disposition attributions that ignores that possibility will misdescribe both natural and engineered systems. The implication reaches beyond idealized wires. Chemical samples are altered by measurement; biological preparations can adapt to stress; agents may learn or lose an ability while being evaluated. The philosophical structure is shared even though the mechanisms differ.

Masks, Antidotes, and Mimics: The Ecology of Manifestation

Alexander Bird's “Dispositions and Antidotes” (1998) challenges Lewis's revision by distinguishing a fink from an antidote. An antidote prevents the characteristic manifestation while leaving the disposition and its immediate basis intact. Suppose a poison retains its toxic properties but an additional substance blocks the physiological pathway by which its effect would be produced. A counterfactual that preserves the poison's causal basis may still fail to yield the expected manifestation. Similarly, a fragile item may be supported or protected in ways that prevent breaking without abolishing its fragility.

A mask or interference factor generally obstructs manifestation without eliminating the power. What exactly counts as an antidote rather than another kind of mask depends on the causal model. The distinction matters less as an exhaustive taxonomy than as a demonstration that the persistence of the basis and the appearance of the manifestation can come apart. The bearer can have the same structural disposition while its surroundings alter the result (Bird, 1998).

A mimic presents the reverse epistemic hazard. A process can produce the expected manifestation even though the original object lacks the power being ascribed. A glass might break because of an unseen external explosion rather than because of the moderate impact being tested. A person might happen to give the right answer by guessing without having the relevant skill. The resulting event cannot, on its own, settle which causal capacity was present. The metaphysical ascription must distinguish an outcome from its explanation.

These cases expose a limitation of crude stimulus–manifestation pairs. In actual systems, manifestation depends on a causal field: structures of the bearer, interactions among participating entities, timing, thresholds, environmental variables, and countervailing processes. The appropriate counterfactual is seldom obtained by adding one stimulus to a world while leaving all the rest unanalyzed. Controlled experiments are powerful precisely because they take extensive pains to stabilize or vary this field. Their sophistication is evidence for the complexity of the target, not for the dispensability of capacities.

The resulting debate includes two recognizable strategies. One preserves a conditional account by enriching the antecedent with explicit favorable conditions, persistence clauses, and causal qualifications. The other concludes that no independently specified counterfactual conditions will analyze dispositionality in general, and takes counterfactual behavior as evidence of a prior capacity rather than as its definition (Mumford, 1998; Molnar, 1999). The first strategy must avoid making the antecedent equivalent to “conditions under which the disposition manifests.” The second must explain the capacity's identity and evidence without presupposing a mysterious primitive whenever an analysis fails.

It is possible to retain a mediated relation between disposition and counterfactual. A disposition licenses a family of context-sensitive counterfactuals, though no single member of the family provides necessary and sufficient conditions for its possession. That family may include probabilistic statements, statements about gradients of stimulation, and statements about characteristic but defeasible outcomes. This position gives a serious role to the conditional approach while recognizing that a real capacity operates among other real capacities.

Causal Bases, Realization, and Intrinsic or Extrinsic Power

A further controversy asks whether every disposition must have a causal basis other than itself. A material's fragility may be explained by bonding, defects, and stress distributions. In many cases the explanatory basis is described with scientific precision, and a disposition ascription summarizes the behavior enabled by that structure. From this it is tempting to infer that dispositions are derivative properties of a fundamentally categorical base. Such an inference requires additional premises, however. The existence of a microstructural explanation does not reveal whether the microstructural properties are themselves categorical, dispositional, or both.

The regress argument is familiar: if every disposition has a distinct causal basis, and every basis is itself dispositional, explanation seems to proceed indefinitely. Some philosophers conclude that a categorical ground must terminate the chain. But the argument's force depends upon the demand for a distinct basis and on the assumption that explanation must terminate in an entity that lacks the very feature being explained. Molnar's project allows powers to be fundamental, while Mumford challenges the inference from grounded dispositions to a uniformly categorical ontology (Molnar, 2003; Mumford, 2006). Another response allows powerful qualities whose qualitative and dispositional aspects are inseparable, without multiplying a separate base behind each power (Heil, 2003).

The possibility of extrinsic dispositions also tests accounts that define dispositional properties by intrinsic constitution. Jennifer McKitrick (2003) argues for extrinsic dispositions: a thing's susceptibility or ability can depend on relations it stands in, even when there is no intrinsic change in the thing. A key is disposed to open a specific lock partly by virtue of the arrangement of that lock; an object can be vulnerable to an external mechanism because of its location or social setting. Some of these cases can be redescribed as intrinsic powers plus extrinsic conditions, but that redescription may change the target attribution. The distinction between an object’s generic capacity and its capacity with respect to a determinate partner is philosophically informative.

Consider the relation between a magnet and a metallic object. If the magnetic disposition is specified generically, its bearer may be described largely through an intrinsic physical profile. If the disposition is specified as an ability to affect this particular object at this particular location, relational facts become indispensable. The primary metaphysical question is whether the ascribed property itself is relational or whether an intrinsic power participates in a relational capacity. Generalizing from one case to every disposition assumes in advance an answer to the individuation problem.

A dispositional account can be physically serious without committing itself to a unique basis at every higher level. The same functional capacity may be realized by heterogeneous mechanisms, and the same underlying structure can support several dispositions. Such many-to-many realization complicates a simple identity theory of dispositions and their bases. It invites an explanatory ontology in which an ascription may track a stable capacity at one scale while its mechanisms are decomposed at another. Identifying one with the other requires an argument about persistence, counterfactual profile, and explanatory indispensability, rather than a mere appeal to the possibility of microphysical description.

The lesson is that there are several non-equivalent reduction projects. A semantic reduction replaces disposition vocabulary with categorical vocabulary. An explanatory reduction shows how one capacity is realized by other mechanisms. An ontological reduction identifies the property with a base or denies the independent status of the ascribed property. Success in the second project does not settle the first or the third. Keeping them distinct prevents a recurrent slide from the legitimate scientific discovery of mechanisms to a premature metaphysical dismissal of powers.

Powers, Directedness, and the Identity of Properties

George Molnar's Powers: A Study in Metaphysics (2003) gives a systematic defense of the irreducibility of powers. His approach treats powers as genuine properties of their bearers and emphasizes intentionality-like directedness: a power is directed toward a possible manifestation even when that manifestation is absent. This feature does not entail that the bearer is conscious or represents the outcome. It identifies a structural analogy with aboutness, namely that the characterization of the present power refers beyond the present event. Molnar further explores actuality, independence, and the distinction between active and passive powers.

Directedness raises a question of individuation. Is a power identified by the manifestation toward which it is directed, by the stimulus–manifestation pair, by its causal role across an entire network, or by a nature prior to all these descriptions? A disposition to dissolve and a disposition to react in an acid may overlap in some cases yet differ in others. Two powers may yield the same observable response through distinct processes. If every manifestation has multiple contributing powers, individuation by outcome alone becomes precarious.

The pure-power conception treats fundamental properties as essentially disposed toward particular patterns of activity. The identity of the property is inseparable from its modal role. Bird (2007) offers one of the most influential systematic developments of this position, connecting property essences to the metaphysical necessity of laws. A categoricalist view instead allows a property to retain its identity while being associated with different nomic roles in other possible worlds; the intrinsic quality is, in principle, separable from its causal behavior. The contrast is clearest at the level of fundamental properties, since ordinary macroscopic dispositions can be acknowledged by either camp.

This gives the quiddity problem its importance. Suppose two metaphysically possible worlds contain properties whose complete causal roles are exchanged. If the properties have primitive identities independent of their roles, the worlds may differ despite an apparent identity of all their structural and causal patterns. Critics ask what distinguishes these worlds and how any knower could discover the difference. Advocates of categoricalism can answer that epistemic indistinguishability does not by itself establish metaphysical identity. Dispositionalists respond that causal role may be sufficient for the identity of a fundamental property. The disagreement concerns the admissible grounds of property identity, not merely the practical utility of disposition talk.

A powerful-qualities position attempts to treat the same concrete property as both characterizing what a thing is like and enabling what it can do. Martin and Heil (1999) and Heil (2003) provide influential articulations. The view aims to resist a split between categorical grounds and separate powers, but it faces a demand for positive explanation: in what sense are the qualitative and dispositional aspects identical rather than simply conjoined? If their identity is asserted as primitive, the view must show how this yields a more intelligible account of causal contribution than the two-sided theories it rejects.

The issue can be formulated without pretending that one ontology has already won. An adequate account of fundamental properties must explain identity, manifestation, causal relevance, and modal stability. Pure powers place modal role at the core of identity. Pure categoricalism grounds the role in further facts such as laws. Powerful qualities seek a unitary identity whose explanatory character is both qualitative and productive. Evaluation should compare each theory's cost in primitive commitments and its ability to handle changing, masked, and relational capacities.

Dispositional Essentialism and the Necessity of Natural Law

The claim that a property has a dispositional essence can take several strengths. A weak version maintains that at least some actual properties have essential causal roles. A stronger version identifies all fundamental natural properties by their powers. A still broader pandispositionalism extends the thesis to properties generally. These versions differ over whether geometrical, structural, mathematical, and other properties fall within the same account. A failure of the strongest version would not automatically refute the weakest.

The guiding inference of dispositional essentialism can be reconstructed as follows. A fundamental natural property P has an essential directedness toward manifestation M under suitable circumstances. If P has that nature in every world in which it exists, then worlds containing P cannot assign it a fundamentally incompatible behavior without changing what P is. A law describing P's characteristic behavior accordingly derives its necessity from the essence of P. Bird's Nature's Metaphysics (2007) develops this strategy as an alternative to both Humean regularity accounts and theories in which nomic necessitation is an external relation among universals.

Its philosophical attraction is clear. Ordinary laws are expressed through properties—charge, mass, spin, and others—and one may wonder why those properties should be associated with just these patterns of behavior rather than wholly different ones. If the behavior is essential to the property, this association is not a further unexplained contingency. The proposed explanation moves necessity into the identity of the constituents of nature.

Two qualifications are necessary. First, an essential disposition is not equivalent to a promise of an exceptionless observed manifestation. Interference, interacting powers, thresholds, and probabilistic outcomes can mediate the exercise of the property. Second, the transition from a dispositional essence to a particular mathematical law requires details about the composition, interaction, and quantitative specification of powers. The bare claim that properties are directed toward manifestations does not deliver Maxwell's equations, a probability distribution, or an equation of motion by itself.

A categoricalist can accept much of this explanatory project while rejecting its metaphysical necessity claim. On a law-first approach, properties are what they are independently of the laws assigning their causal roles. On a Humean best-system approach, laws summarize the overall pattern of facts with an optimal balance of simplicity and informational strength. On the Armstrong–Dretske–Tooley family of approaches, laws involve a necessitation relation among universals. The dispositional essentialist must therefore explain not simply why regularity requires explanation, but why an essentialist account is better than its strongest competitors at the relevant explanatory task (Bird, 2007; Armstrong, 1983).

A major objection concerns modal epistemology. If physical science identifies a property through experimentally established manifestations, how does that warrant the stronger assertion that its observed causal profile is essential in all metaphysically possible worlds? Scientific discovery may justify a robust theory of what the property actually does. A further philosophical argument is required to infer what it could not have done while remaining the same property. The essentialist may invoke the theory of property identity, but that makes the result depend on a contested metaphysical criterion.

A second objection concerns the apparent contingency of many scientific generalizations. The emergence of complex structures, initial conditions, and environmental interactions can change what patterns occur, even where the fundamental dispositional essences are assumed constant. A power-based account must distinguish necessary constraints grounded in fundamental properties from contingent outcomes generated by particular arrangements. When it makes this distinction clearly, it can preserve a serious account of chance and historical contingency within a world of essential causal natures.

The most compelling essentialist insight is a demand for the integration of what a property is with what it can contribute. Its largest remaining obligation is to derive, rather than merely redescribe, the explanatory structure of natural law.

Causation as the Exercise and Interaction of Powers

Causation is a natural testing ground for powers ontology. An ordinary causal explanation says that salt dissolved because it was immersed in water under suitable conditions. The explanation appears to mention both a stable capacity and the process in which it was exercised. A power theorist can interpret the event as an interaction in which the causal contributions of participating things explain the outcome. The theory thereby locates causal production in actual properties of the participants.

Mumford and Anjum's Getting Causes from Powers (2011) develops a powers-based theory that treats causal production as complex and context sensitive. The book describes things as possessing clusters of powers and emphasizes that outcomes arise from multiple contributions rather than a solitary, automatically sufficient cause. This connects with J. S. Mill's attention to the total cause and with contemporary scientific practice, where the presence of a causal factor is frequently assessed against a background of enabling and counteracting conditions (Mumford & Anjum, 2011, ch. 1).

Consider a seed's germination. The seed has biological capacities, yet moisture, temperature, oxygen, nutrient availability, and developmental integrity all matter. Saying that the seed has a power to germinate is compatible with its failure to do so under a particular environmental condition. A reductive account may cite internal biochemistry plus background laws; a powers account interprets those biochemical processes themselves as the exercise of lower-level capacities. The two descriptions compete only if they provide incompatible metaphysical explanations. They can coexist as different levels of one causal investigation.

The metaphysics of powers must also face preemption and overdetermination. If two independently sufficient processes are directed toward the same outcome and one acts first, the other may retain its disposition without making the relevant causal contribution. If two processes jointly produce an outcome, each may be causal without being individually sufficient. A simple model that identifies being a cause with manifesting a disposition can misclassify these cases unless the conditions of contribution and the individuation of manifestations are specified.

Causal processes can be probabilistic. A carcinogenic exposure may raise the likelihood of disease without producing it in any particular case. The power to increase a probability should not be identified with a deterministic power that fails on some occasions. Dispositional theorists can construe propensities or tendencies as genuine directed capacities, but must indicate whether the probability is objective, epistemic, population-relative, or generated by underlying indeterminism. Without that distinction, attributing a probabilistic power may merely redescribe an observed frequency.

There is a residual circularity challenge. If a power is defined as what produces an effect, while production is defined as the exercise of a power, the combined theory explains neither term through an independent reduction. It may still offer a legitimate metaphysical account if power and production are taken as jointly primitive or if their coherence is secured by broader constraints. But the theory must acknowledge the primitive and show its explanatory usefulness. The strongest causal-power account is one that identifies actual causal organization, explains failed and successful manifestations, and tells us why particular experiments discriminate among alternative causal hypotheses.

The relation between cause and power is thus close without being conceptually identical. Powers are persistent resources for possible causal participation; causes are particular contributions to actual processes or differences made in particular circumstances. Keeping the two distinct helps explain how an unexercised power is real and how an exercised power may interact with other causal factors.

Potentiality and Metaphysical Modality

A disposition is modal: it bears on what can occur, not only on what has occurred. But the reach of this modal content is debated. Possibility may be analyzed using possible worlds taken as primitive, as in traditional modal semantics, or it may be grounded in the potentialities of actual things. Vetter's Potentiality: From Dispositions to Modality (2015) develops a sustained version of the latter approach. It proposes a general metaphysics of potentiality broad enough to include ordinary dispositions, marginal abilities, joint potentialities, and iterated potentialities, and aims to connect these to metaphysical possibility.

The attraction of this approach is its actualist grounding. To say that something is possible is to trace that possibility to what actual entities can be or can become, perhaps through iterated changes. One need not begin with a vast inventory of independently existing worlds. A capacity of a thing is the starting point for constructing modal truth, rather than a property described only after all possible worlds are specified.

Yet three difficulties require distinct treatment. First is extensional adequacy: do the potentialities of actual things encompass all genuine metaphysical possibilities, including possibilities concerning beings never actualized? Second is formal adequacy: do the principles governing potentiality validate the logical relations that metaphysical possibility and necessity are supposed to satisfy? Third is semantic adequacy: can the theory explain ordinary modal language without constantly shifting between practical, nomological, and metaphysical possibility? Vetter explicitly organizes her defense around these kinds of demands (Vetter, 2015, especially chs. 4–6).

The status of apparently impossible manifestations also matters. A badly deformed key might have almost no chance of opening any actual lock but still possess some minimal structural potentiality under highly unusual circumstances. On the other hand, no degree of ordinary material capacity appears to make contradictions genuinely realizable. A theory must discriminate weak potentiality, extremely remote possibility, and impossibility rather than relying on an undifferentiated “could.”

There is also a temporal problem. Capacities change, emerge, and disappear. An organism that cannot presently perform a behavior might acquire the capacity by development or training. Iterated potentialities aim to capture possibilities of acquiring further possibilities. This move is powerful, but it invites a question about the limits of iteration: if every imaginable transformation is admitted without constraint, the grounding theory risks becoming extensionally indiscriminate. The present constitution of actual things must place substantive restrictions on accessible modal paths if potentiality is to explain possibility.

A further comparison with Aristotle is clarifying. Aristotle's account of dunamis unfolds inside a metaphysics of substances, natural forms, and actuality, whereas Vetter's program aims to ground modality more generally through potentialities. Both recognize a relation between what exists and what it could do or become; their accounts of which possibilities are grounded, and how, belong to different philosophical systems. Historical analogy supplies a question, not an identity of doctrines.

The strongest conclusion is modest but significant. Dispositions provide indispensable local modal structure for scientific and ordinary explanation. Whether all metaphysical possibility is reducible to actual potentialities remains an open substantive thesis, on which the full ontology and logic of possibility depend.

Scientific Capacities, Measurement, and Explanatory Models

Scientific theories regularly assign capacities to systems whose characteristic outcomes are observable only under carefully arranged conditions. The same material can be conductive in one state and insulating in another, a drug can exhibit efficacy in one population and little benefit in another, and ecological systems can respond differently to apparently similar disturbances. These are reasons to improve the specification of dispositions and their enabling conditions, not automatically to eliminate them.

Nancy Cartwright's Nature's Capacities and Their Measurement (1989) argues that capacities play an essential role in scientific causal inference and measurement. She distinguishes general causal claims about what a factor can produce from particular regularities observed in selected populations or arrangements. Her case studies, including econometrics and quantum mechanics, support the thesis that causal knowledge often travels through recognition of relatively stable capacities across varying contexts, though the conditions under which a capacity is expressed can change (Cartwright, 1989, introduction and ch. 4).

A capacity is often inferred through an experimental contrast. When an intervention changes an outcome while relevant variables are controlled, investigators gain evidence for a factor's causal role. This inference is not equivalent to direct observation of an isolated power. It is mediated by models, assumptions about confounding, measurement reliability, and the stability of the tested system. A metaphysical theory earns scientific relevance by clarifying these procedures without pretending to replace them.

The concept of a causal basis is particularly useful for connecting science and metaphysics. In chemical solubility, bonding and thermodynamic conditions can furnish an explanation. In material fracture, microcracks, stress concentrations, and temperature bear on manifestation. In biological regulation, molecular mechanisms and the organization of whole organisms jointly determine what a capacity can produce. An ontology of dispositions must be compatible with the possibility that new evidence revises the identified basis, the boundary of the relevant bearer, or the expected manifestation profile.

Science also reveals why powers should not be equated with deterministic laws. Models can describe tendencies that contribute to outcomes in the presence of countervailing processes. In epidemiology, economics, and ecology, a causal factor may exert a stable influence while observed effects vary. The philosophical task is to say when this stability is genuine and what evidential grounds distinguish it from mere statistical association. The invocation of a capacity should add testable explanatory commitments: specify the factor, a class of interventions, expected changes, and plausible interfering mechanisms.

The relation between disposition talk and scientific realism remains open to competing interpretations. A realist can understand successful capacity attributions as referring to genuinely existing causal characteristics. A constructive empiricist or instrumentalist may regard them as valuable elements of model-based prediction without accepting their literal metaphysical interpretation. The practical reliability of an attribution does not settle the entire ontological dispute. Conversely, an unresolved metaphysical debate does not deprive the attribution of empirical content.

The best scientific application therefore combines ontology with methodological restraint. It treats powers as candidate constituents of explanation and submits the corresponding claims to causal analysis, predictive use, intervention, and revision. It resists the temptation to infer a full metaphysical theory from isolated examples of successful scientific vocabulary.

Dispositions, Probability, and the Problem of Multi-Track Powers

Many familiar dispositions are multi-track: they can manifest in more than one way. Elasticity includes resistance to deformation and recovery of shape, a biological immune capacity may generate several response pathways, and a skilled agent may perform a task through numerous techniques. Attempting to individuate each as a unique stimulus–manifestation pair may fragment a recognizable capacity into a vast family of artificially separate properties.

A multi-track conception understands the disposition through a structured profile of responses over relevant ranges of circumstances. The profile may be continuous or threshold dependent, and one manifestation can involve both a process and a resulting condition. Such a profile is typically more faithful to scientific work, where a capacity is characterized by a response curve, mechanism, or ensemble of measurements rather than a single yes-or-no test. Yet the metaphysical problem remains: which of these response characteristics constitute the power's identity, and which are contingent consequences of the environment?

The issue becomes sharper when an attribution is gradable. A more fragile glass is likely to break under weaker stresses or across a broader range of perturbations. Manley and Wasserman (2008) emphasize that philosophical semantics should attend to dispositional predicates admitting degrees. A theory that identifies fragility with the result of one specified test fails to explain the comparative judgment unless it includes thresholds, probability, or a space of relevant counterfactual situations.

Propensity interpretations of chance bring a related difficulty. An objectively probabilistic power is often taken to support a particular chance of manifestation. But the relevant propensity may attach to the object, to an experimental arrangement, or to an entire causal setup. An unstable nucleus's propensity to decay and a coin's propensity to land heads do not obviously instantiate the same kind of fundamental property. Some probability distributions arise from dynamics, some from statistical modeling of complex deterministic processes, and some may be taken as irreducibly stochastic. A powers theory should not legislate one metaphysical explanation for all three through the ambiguous word “tendency.”

Where a probability is genuinely connected to a power, the relation may be expressed as a constrained profile: under a specified range of circumstances, exercise of this power contributes a characteristic change in the distribution of outcomes. That formulation does not presuppose that a probability belongs to the bearer independently of context. It also allows multiple interacting capacities to determine the final distribution without treating every outcome as a failed or successful manifestation in a binary sense.

The resulting philosophical account gains precision when dispositions are represented as dynamically structured and multiply realizable. Whether these structures ultimately reduce to categorical grounds, essentially powerful properties, or powerful qualities is a separate level of inquiry. The existence of gradable and multi-track dispositions is a constraint on all three theories, not a decisive victory for any one of them.

Dispositions in Mind, Agency, and Living Systems

The metaphysics of dispositions enters debates about mind and agency because abilities, beliefs, intentions, character traits, and skills commonly outlast their episodes of exercise. A person who knows how to swim remains able to swim between occasions of entering water; someone who believes a proposition need not be consciously entertaining it at every moment. These are paradigms of persistence without manifestation. Their structure is philosophically continuous with physical dispositions in some respects, but it also introduces normativity, social context, learning, and the distinction between competence and performance.

Ryle's treatment of knowing how resists the reduction of intelligent practice to a sequence of internally rehearsed propositions. The ability is displayed in appropriately organized performances and in a person's responsiveness to varying circumstances (Ryle, 1949, ch. 2). This helps separate possessing a skill from merely succeeding on one occasion. A lucky novice may imitate an expert's result without possessing the capacity to do so reliably. Conversely, an expert may fail because of injury, distraction, or hostile conditions while retaining a more general skill. These examples reproduce the general structure of mimics, masks, and context-sensitive manifestation, enriched by the standards according to which a performance counts as competent.

Agency introduces a further dimension: the power to do something may involve the agent's reasons and opportunities rather than only a mechanical stimulus. The ability to speak is not normally exercised whenever a relevant external stimulus is present; an agent can decide whether, when, and how to use it. The Aristotelian distinction between rational and nonrational capacities remains useful here, provided the historical formulation is not mistaken for a complete contemporary theory of intentional action. The analysis of practical capacities belongs in dialogue with the philosophy of action, while the general ontology of dispositions concerns what makes their possession possible.

In biology, capacities are often organizational and historical. An organism's ability to regenerate tissue, resist disease, learn, or reproduce is grounded in complex interacting mechanisms. It can depend upon developmental stage, ecological context, or relations with other organisms. A purely intrinsic characterization may omit essential enabling systems. At the same time, a purely environmental characterization overlooks the organization that allows a particular organism to respond. Joint and extrinsic potentialities become especially important when describing symbiotic systems, populations, or institutional agency (Vetter, 2015, ch. 4).

Artificially constructed systems present another legitimate application. A machine's capacity to recognize a pattern, actuate a component, or transform information depends on hardware, software, training history, energy, maintenance, interfaces, and environmental compatibility. Such an attribution can be investigated through the same bearer–basis–manifestation–context questions used elsewhere. Whether a particular artificial system possesses additional forms of agency, normativity, or understanding is a distinct philosophical inquiry. The ordinary metaphysics of powers does not settle those questions through the mere presence of sophisticated performance.

These examples establish a disciplined extension of the concept. The word “power” can express causal potential, skill, social authority, and political domination; their meanings overlap only partially. A person's legal authority is institutionally constituted, while electrical conductivity is a physical characteristic. It is legitimate to study both as conditional capacities, but their criteria of possession, grounds, and manifestations differ. Philosophical generalization should preserve those differences.

Critical Evaluation: What Powers Theories Explain and What They Assume

A rigorous assessment must test the strongest available version of each family of accounts. Conditional theories offer a way to connect disposition predicates with observable and counterfactual behavior. Their challenge is to specify the relevant circumstances without excluding precisely the interference cases that make ordinary dispositions useful. A simple conditional is often false, while a suitably enriched conditional risks circularity or unmanageable complexity. The graded and proportional treatment proposed by Manley and Wasserman (2008) improves the account of comparisons and contextual variations but raises its own questions about which possible cases matter and how they should be weighted.

Categoricalist theories preserve a distinction between a property's identity and its causal role. Their attraction includes compatibility with a Humean picture in which fundamental facts are arranged without primitive modal necessity and with scientific accounts that decompose capacities into structures. Their burden is to explain the modal truth of disposition ascriptions and the coherence of property identities across worlds whose causal profiles are altered. It is possible to accept primitive quiddities, but doing so carries an explanatory price when those identities appear inaccessible to every possible pattern of causal observation.

Pure-power theories identify fundamental properties through their directedness and make causal roles constitutive of identity. They therefore supply a direct response to quiddistic skepticism and a possible ground for necessary natural laws. Their burden is to explain how directedness is individuated, how relations among powers yield definite manifestations, and why the resulting system has the correct modal structure. A power attributed simply because a manifestation occurs does not advance explanation. The theory must specify a stable causal architecture and distinguish counterfactual predictions from retrospective labels.

Powerful-qualities theories propose that a property is simultaneously a way its bearer is and a source of possible activity. The attraction is ontological economy: the qualitative and dispositional descriptions do not require two properties occupying one place. The problem is whether identifying the aspects yields an informative account rather than a terminological resolution. Their advocates must clarify the sense in which qualitative identity constrains manifestation and why that constraint is neither an external law nor an unexplained postulate (Heil, 2003; Martin & Heil, 1999).

A central criticism of disposition-based explanation is illustrated by the familiar charge of a “dormitive virtue.” If one says that a substance induces sleep because it has the power to induce sleep, no new mechanism or discriminating condition has been identified. This objection is decisive against explanatory circularity but does not render every disposition ascription trivial. When a capacity is independently characterized through causal structure, comparative experiments, response curves, and different classes of manifestation, it can contribute to explanation. The decisive distinction is between explanatory use and mere relabeling of an effect.

Power theories also confront a question about levels. Is the power of a molecule to participate in a reaction fundamentally different from the power of an organism to survive an environmental challenge? Some philosophers seek ontological uniformity by explaining all higher-level dispositions through fundamental powers. Others accept higher-level capacities as real because they support robust counterfactual generalizations and interventions. The issue is not resolved by the word “emergence.” Its resolution depends on whether higher-level properties add causal distinctions, whether their realizations are stable across mechanisms, and what sort of metaphysical dependence is being defended.

Finally, the epistemic asymmetry between powers and manifestations is unavoidable. Observers see particular results under particular conditions and infer capacities that may exceed what was tested. Scientific inquiry refines this inference through replication, intervention, mechanistic investigation, and explicit uncertainty. Metaphysics supplies interpretations of that practice. When an interpretation claims that the essence of a property determines what must occur across every possible world, it must offer a bridge from the available evidence to the broader modal conclusion. Assertions of necessity do not themselves constitute that bridge.

A Systematic Reconstruction: Six Explanatory Relations

The preceding historical and analytical chapters permit a synthesis organized around six relations. The first is the bearer relation: a disposition is attributed to a thing, system, process, or organized plurality. The second is the grounding relation: the attribution has a basis in the bearer’s constitution, its relations, or a combination of both. The third is the directedness relation: the property is intelligible through one or more possible manifestations. The fourth is the conditioning relation: different circumstances alter the range of ways the disposition can be exercised. The fifth is the interaction relation: the bearer’s contribution combines with other capacities and constraints in producing or preventing outcomes. The sixth is the modal relation: the attribution supports claims about possible and expected manifestations that outrun actual instances.

This reconstruction is an analytical framework for the subject, not a newly proposed canonical metaphysical theory. It permits competing ontologies to specify where their disagreements fall. The categoricalist may take the grounding relation to terminate in nondispositional properties and understand the modal relation through laws or the total distribution of facts. The pure-power theorist may regard the grounding and directedness relations as essentially connected at the fundamental level. The powerful-qualities theorist may seek a unitary property that grounds its own manifestation profile. The framework clarifies the rival positions without deciding the contest by definition.

The framework has four significant consequences. First, unmanifested dispositions are intelligible as present facts with modal profiles. A power can remain actual as a property without its manifestation becoming an actual event. Second, manifestation is relationally situated: neither an isolated stimulus nor an isolated basis exhausts the causal field. Third, there is no general inference from semantic irreducibility to ontological fundamentality. The failure of a conditional definition creates work for semantics, while a case for basic powers requires a theory of properties and causal explanation. Fourth, modal scope must be calibrated. What a thing can ordinarily do, what laws permit, and what is metaphysically possible can stand in complex relations without being interchangeable.

The reconstruction also yields a criterion for substantive explanatory progress. A successful disposition attribution should discriminate between otherwise plausible hypotheses about a bearer, specify actual or possible interactions, and support further investigation. For a fundamental theory, this can involve distinguishing competing candidate property identities and deriving constraints on lawlike behavior. For an applied causal model, it can involve identifying enabling conditions or predicting results of interventions. Where the attribution performs none of these tasks, it remains an uninformative label.

There is a further consequence for the philosophy of knowledge. Disposition attributions express knowledge of possibilities grounded in actual structures or relations. This knowledge is corrigible because new circumstances can reveal finks, masks, unexpected pathways, or different causal bases. The inferential leap from observed manifestations to enduring powers is legitimate when sustained by systematic causal evidence; it remains open to revision. The intelligibility of that practice does not depend on choosing one final ontology in advance.

Open Philosophical Questions

The first open issue is the individuation of fundamental powers. If properties are identified by causal roles, the theory must say which roles are essential, whether the same manifestation may have distinct bases, and how a power's identity persists through changes in interaction. The issue is especially difficult for multi-track powers and properties represented by symmetries or mathematical structures in physics.

The second concerns the scope of dispositional essentialism. It remains unsettled whether all fundamental natural properties are powerful, whether some are categorical, and whether relational or geometrical properties need the same explanation as dynamic properties. An ontology that classifies all properties as powers must confront apparently structural features whose causal directedness is not evident.

The third concerns laws, chance, and necessity. If laws derive from property essences, a philosophical account should show how quantitative laws and probabilistic dynamics follow from those essences. The general thesis is suggestive, but without a derivation it may not explain the detailed form of scientific law.

The fourth concerns emergence and level dependence. Higher-level dispositions can have stable counterfactual profiles even when realized by multiple lower-level mechanisms. Determining whether this warrants emergent causal powers requires a defensible account of dependence, realization, intervention, and causal exclusion.

The fifth is epistemological. Experimental access to a disposition is always mediated by circumstances that may affect the property being investigated. The question is how to distinguish the power possessed from the power manifested, and how to justify the modal reach of a scientific attribution under uncertainty.

The sixth concerns practical agency and artificial capacities. As engineered systems acquire complex and adaptive repertoires, the concepts of ability, competence, agency, and norm-guided action need precise boundaries. General dispositional metaphysics provides the foundational vocabulary; determining which further concepts apply requires research into the actual organization and activity of the systems concerned.

Philosophical Synthesis

The historical trajectory from Aristotle through the early modern controversy and the contemporary literature discloses one persistent problem in several changing forms. Things are described in terms of what they can do; those descriptions remain relevant when the corresponding events do not occur. Aristotle embedded the distinction in an account of actuality, change, and substance. Locke and Leibniz reshaped it through competing understandings of passive and active power. Hume demanded an account of the idea and warrant of necessary connection. Twentieth-century conditional analyses sought to clarify the inferential content of disposition talk, and the objections from finks and antidotes revealed the importance of temporal change and causal context.

Contemporary property metaphysics turns these lessons into competing theses about what the world contains. Categoricalism separates property identity from causal role; dispositional essentialism makes the role essential to identity; powerful qualities seek their identity in a single property's qualitative and productive character. Each theory preserves an insight and incurs a cost. The modern dispute has advanced through the refinement of these costs, particularly the treatment of modal necessity, causal explanation, and the individuation of properties.

The analysis supports a robust realist treatment of many ordinary and scientific capacities: their existence is supported by organized mechanisms, stable intervention profiles, and the explanatory success of attributions that remain true without current manifestation. This commitment does not entail that every capacity is a fundamental property. Fundamental powers remain a possible metaphysical ground; their strongest defense would derive explanatory and modal constraints that categoricalist competitors cannot supply at comparable theoretical cost.

The six-relation reconstruction locates the unity of the subject. A disposition is understood through its bearer, ground, directedness, conditioning, interactions, and modal consequences. The relations are analytically distinguishable and frequently empirically connected. Their organization explains why a disposition can be actual yet unexercised, why a trigger can fail to produce a manifestation, and why scientific capacities can be informative without being metaphysically primitive. It also explains why reductions of the entire subject to a single conditional repeatedly fail: the target is a structured explanatory relation rather than one uniform pattern of hypothetical behavior.

Conclusion

Dispositions and powers belong at the intersection of metaphysics, philosophy of science, modality, and the explanation of action. Their central philosophical significance is the reality and intelligibility of capacities that reach beyond their occasions of exercise. The distinction between possession and manifestation is established across classical and contemporary arguments; the nature of the properties making it true remains a live metaphysical controversy.

Historical reconstruction establishes the enduring importance of potentiality and activity while showing how substantially their theories have changed. Argumentative analysis shows why the simple conditional account fails in the presence of finks, antidotes, masks, and mimics, and why sophisticated analyses must confront graded, relational, and changing capacities. Comparative metaphysics distinguishes categorical grounds, irreducible powers, powerful qualities, and explanatory accounts of causation and natural law.

The resulting philosophical position is that dispositional attributions require a structured account of bearer, ground, possible manifestations, circumstances, interaction, and modal scope. They can provide indispensable explanatory knowledge without a presupposition that their ultimate ontology has already been resolved. What remains at issue is the metaphysical explanation of that knowledge: whether nature's most basic properties are constituted by their powers, governed by laws independently of those powers, or understood through a unity of qualitative and dispositional being. That question is the continuing philosophical center of the research.