- Atlas
- Philosophy
- Philosophy Of Causation
Philosophy Of Causation
Causation is the relation that connects one event to another when the first brings about, produces, or is responsible for the second.
12 frameworks, from 350 BCE to today · Metaphysics · Philosophy
- Four causes 350 BCE-1600
- Mechanistic Causation 1600-1750
- Occasionalism 1650-1720
- Humean Regularity Theory 1739-1970
- Humean skepticism about necessary connection 1739-Present
- Kantian Transcendental Causation 1781-1850
- Probabilistic Causation 1956-Present
- Causal Powers and Dispositionalism 1970-Present
- Counterfactual Theory 1973-Present
- Causal Pluralism 1980-Present
- Causal model 2000-Present
- Interventionist and manipulationist accounts 2000-Present
Overview
Causation is the relation that connects one event to another when the first brings about, produces, or is responsible for the second1. Striking a match causes it to light; smoking causes cancer; a drought causes a famine1. Everyone uses the notion, and almost no one can say precisely what it is. The philosophy of causation is the attempt to say what it is: whether causes make their effects happen or are merely regularly followed by them, whether causal claims are about what would have happened otherwise, whether causation is one thing or many, and how we can infer causes from data. The stakes are practical as well as theoretical. Science, medicine, law, and moral responsibility all lean on causal claims, and a theory of causation that makes those claims unintelligible—or too easy—changes what we are entitled to say and do.
Aristotle's Four Causes and the Pre-Modern Inheritance
Aristotle (4th century BCE) gave causation its first systematic treatment, and his framework dominated for nearly two thousand years. He distinguished four ways of answering "why?"—later called the four causes12. The material cause is what something is made of (bronze for a statue); the formal cause is its defining structure or shape; the efficient cause is the source of change (the sculptor); and the final cause is the end or purpose for which it exists1. For Aristotle, explaining a thing meant giving all four, and the final cause was not a curiosity but central: natural things have ends built into them, so acorns are for oaks.
This framework was absorbed into Islamic and Christian scholastic philosophy3. Avicenna and Averroes elaborated it; Thomas Aquinas integrated it into Christian theology in the 13th century. The efficient cause—the maker or agent—gradually became the focus, but the final cause retained its authority in accounts of nature and human action. The framework's strength was its breadth: it explained artifacts, organisms, and human purposes in one scheme. Its weakness, which became visible in the Scientific Revolution, was that it treated purpose as a real feature of the natural world rather than something imposed by minds.
Mechanism, Occasionalism, and the Attack on Final Causes
The Scientific Revolution of the 16th and 17th centuries dismantled the Aristotelian picture. Galileo, Descartes, and Newton described nature as matter in motion governed by laws, and purpose dropped out of physics. Mechanistic causation held that causal interaction is the transfer of motion between bodies by contact or impact—a billiard-ball model. Descartes (1596–1650) made this sharp: bodies are extended substance, and all causation between them is mechanical push and shove.
But mechanism created a problem. If mind and body are distinct substances, how can the mind cause the body to move, or the body cause sensations? Descartes' follower Arnold Geulincx and, more famously, Nicolas Malebranche (1638–1715) answered with occasionalism: created things have no genuine causal power of their own. God is the sole true cause; when a billiard ball strikes another, the occasion is merely the occasion on which God produces the motion4. Malebranche's argument was that no necessary connection can be found between any two created things—only God's will makes the effect follow. Occasionalism preserved divine omnipotence and the mechanical picture at once, but at the cost of making all natural causation a kind of divine habit.
Hume: Causation as Constant Conjunction
David Hume (1711–1776) turned the occasionalist observation into a general empiricist challenge. Examining what we actually observe when one billiard ball strikes another, Hume found no impression of a "necessary connection" passing between them—only one event followed by another, repeatedly5. From this he drew his famous analysis: the idea of causation reduces to three elements—contiguity in space and time, priority of cause to effect, and constant conjunction, the regular pairing of like causes with like effects. The felt "power" or "necessity" is not in the world but in the mind, a habit of expectation built up by repetition15.
Hume's skepticism about necessary connection set the agenda for everything after. If causation is just regular succession, what distinguishes a genuine causal sequence from a mere correlation—night following day? And how can we justify inferences from past regularities to future ones? Hume's own answer, that such inferences rest on custom rather than reason, was as troubling as the analysis itself5. The problem of distinguishing causation from correlation, and of grounding induction, has never fully gone away.
Kant's Transcendental Response
Immanuel Kant (1724–1804) answered Hume by relocating necessity. In the Critique of Pure Reason (1781), Kant argued that causation is not a feature we read off experience but a category the mind imposes on it67. For experience to be possible at all—for us to have an objective world of enduring objects undergoing change—the understanding must apply the concept of cause and effect. Transcendental causation is thus a condition of experience rather than an empirical discovery. Kant's move preserved the necessity Hume denied, but at the price of making causation a structure of our cognition rather than of things in themselves. Later German idealists, notably Hegel, developed causation dialectically, but the Kantian settlement—causation as a precondition of objective experience—remained a live option, and it resurfaces in contemporary work on causal explanation and the "interventionist" idea that causal claims are about what would happen under manipulations of a system.
Regularity Theory and Its Refinement
In the 19th and early 20th centuries, empiricists tried to make Hume's analysis rigorous. John Stuart Mill (1806–1873) formulated methods of causal inference—agreement, difference, concomitant variation—that are still taught as Mill's Methods. The regularity theory proper is associated with logical empiricists: causation is nothing over and above instantiations of laws, and laws are universal regularities. A cause is an event of a type always followed by an event of another type.
The theory's appeal is ontological austerity: no mysterious powers, just patterns. Its problems are equally clear. Regularities can be accidental (every coin in my pocket is silver, but that is not a law), and causes can be singular (a particular fall broke a particular arm, even if no universal law covers exactly that pair). Moreover, regularity theory struggles with asymmetry: the height of a barometer is regularly followed by storms, but the barometer does not cause them. Attempts to repair the theory—requiring laws to be part of a systematic deductive network, as in the deductive-nomological model of explanation—kept the empiricist spirit but never fully solved the problem of spurious regularities.
Probabilistic Causation
Much of the world's causation is probabilistic: smoking raises the probability of cancer without guaranteeing it. Probabilistic causation, developed from the mid-20th century by philosophers such as Hans Reichenbach and later Patrick Suppes and Wesley Salmon, holds that a cause raises the probability of its effect. Reichenbach's The Direction of Time (1956) introduced the idea that causes screen off their effects—once you know the cause, intervening factors become irrelevant—and used this to define causal direction. The approach made causation scientifically respectable in contexts where determinism fails, and it connected causation to statistics.
But probability-raising is not sufficient: a falling barometer raises the probability of rain without causing it. The core difficulty is screening off and common causes—two effects of a shared cause are correlated without either causing the other. Probabilistic theories had to build in conditions excluding such spurious correlations, and the resulting machinery fed directly into the next development.
Counterfactuals and the Possible
The most influential 20th-century turn was the counterfactual theory, given its classic statement by David Lewis in "Causation" (1973). Lewis proposed that event c causes event e just in case, had c not occurred, e would not have occurred. The analysis uses the possible-worlds semantics for counterfactuals that Lewis had developed: "if c had not happened, e would not have happened" is true when the nearest possible world where c fails is a world where e also fails. This captured the intuition that causes make a difference—remove the cause, and the effect goes with it—and it handled singular causation and asymmetry better than regularity theory.
Lewis later refined the account to handle preemption (a backup cause ready to step in) and overdetermination (two causes each sufficient for the effect), and the counterfactual approach became the dominant framework in analytic metaphysics. Its difficulties are technical but deep: the semantics of counterfactuals is sensitive to how we describe events, and "nearest world" comparisons can be gerrymandered. Still, the idea that causation is difference-making—that causes are the things whose absence would have changed the outcome—remains the intuitive core of much contemporary work.
Powers, Dispositions, and the Revival of Real Necessity
Against the Humean consensus, a minority tradition insisted that causation is a real, productive relation. Causal powers and dispositionalism hold that things have intrinsic dispositions—fragility, solubility, magnetism—that are genuinely causal, not mere regularities. A glass is fragile because of its molecular structure, and that structure makes it break when struck. This tradition, associated in recent decades with Rom Harré, Roy Bhaskar, and later Stephen Mumford and Rani Lill Anjum, draws on Aristotelian and scholastic ideas and on the intuition that science discovers mechanisms, not just correlations. Its claim is that regularity theories get the order of explanation backwards: things are regularly followed by their effects because they have powers, not the other way around.
The debate between Humeans and powers theorists is one of the liveliest in contemporary metaphysics. Powers theorists point to fundamental physics, where dispositions like charge and mass seem irreducible; Humeans reply that positing powers explains nothing that regularities cannot, and that powers are metaphysically obscure. The dispute turns on deep questions about what laws of nature are and whether necessity can be observed.
Interventionism and Manipulation
A parallel development, rooted in statistics and computer science as much as philosophy, produced the interventionist or manipulationist account. The core idea, articulated by Judea Pearl in Causality (2000) and by James Woodward in Making Things Happen (2003), is that c causes e if an intervention on c—a manipulation that changes c while leaving other influences on e fixed—would change e. This operationalizes causation: causal claims are claims about what would happen under controlled change. Pearl's formal apparatus, causal models and inference formalisms (directed acyclic graphs, do-calculus), gave researchers a rigorous way to represent causal structure and to infer causes from observational data, and it has been enormously influential in epidemiology, economics, and machine learning. Woodward's version ties causation to invariance under intervention and to explanation: a causal explanation is one that tells you what would happen if you intervened.
Interventionism inherits the counterfactual spirit but makes it tractable. It faces questions about interventions that cannot be performed (cosmological events, past causes) and about whether the account is circular, since interventions are themselves causal notions. Its defenders argue that the circularity is benign and that the account captures what scientists actually do when they identify causes.
Pluralism and the Present Landscape
Running through all of this is the question of whether causation is one relation or many. Causal pluralism—defended by philosophers such as Nancy Cartwright and, in different forms, by Christopher Hitchcock and others—holds that different causal concepts serve different purposes: physics may need productive powers, medicine difference-making, law responsibility, and statistics probabilistic dependence. On this view, no single analysis will capture every legitimate use of "cause," and the search for one true theory is misguided.
Today the field is divided but productive. The counterfactual and interventionist traditions dominate analytic metaphysics and the philosophy of science, increasingly formalized with causal models and graph theory. Powers and dispositionalist accounts remain a serious rival, especially among philosophers of physics and biology. Probabilistic and regularity approaches survive in statistics and in the philosophy of induction. Pluralism is a growing minority position. The open questions are old ones in new dress: What grounds causal asymmetry1? Can causation be reduced to non-causal facts, or is it fundamental? How do causal claims in the special sciences relate to those in physics? And what, exactly, licenses the inference from correlation to cause—the question Hume raised, which no subsequent theory has fully laid to rest.
References
- Causality (introduction, Probabilistic causation, Aristotelian, Derivation theories), Wikipedia
- Four causes, Wikipedia
- Aristotle, Wikipedia
- Occasionalism, Wikipedia
- David Hume, Wikipedia
- Transcendental idealism, Wikipedia
- Immanuel Kant, Wikipedia
The frameworks of Philosophy Of Causation
Each framework is a way of working the field has used: when it held, and the question it answers.
- 350 BCE-1600
- 1600-1750
- 1650-1720
- 1739-1970
Humean Regularity Theory
What do we observe when one event causes another?
- 1739-Present
Humean skepticism about necessary connection
Is there any observable necessary connection between events?
- 1781-1850
Kantian Transcendental Causation
How is causal necessity possible in experience?
- 1956-Present
Probabilistic Causation
How can a cause raise the probability of its effect?
- 1970-Present
Causal Powers and Dispositionalism
Do things have real powers to produce effects?
- 1973-Present
- 1980-Present
- 2000-Present
- 2000-Present
Interventionist and manipulationist accounts
What would happen if we intervened to change something?
How the ideas connect
Each arrow runs from an earlier idea to the later one it shaped. The list below says what passed between them.
Mechanistic Causation reacted against Four causes
Mechanistic causation arose in the Scientific Revolution as a rejection of Aristotelian formal and final causes, replacing them with matter in motion and contact impact. Galileo and Descartes defined the new physics against the scholastic four-cause framework.
Carried by Galileo Galilei; René Descartes.
Occasionalism grew out of Mechanistic Causation
Occasionalism grew directly out of Cartesian mechanism: if bodies are purely extended and interact only by contact, mind-body and body-body causation become unintelligible, so Malebranche and Geulincx concluded that God is the sole true cause.
Carried by Nicolas Malebranche; Arnold Geulincx.
Humean Regularity Theory influenced by Occasionalism
Hume took over the occasionalist observation that no necessary connection is observable between created things and turned it into a general empiricist analysis of causation as constant conjunction. Malebranche's argument that only God's will links events became Hume's claim that the link is a habit of the mind.
Carried by David Hume; Nicolas Malebranche.
Kantian Transcendental Causation reacted against Humean skepticism about necessary connection
Kant defined his transcendental deduction against Hume's skeptical conclusion that causal necessity is merely a habit of expectation, relocating necessity from the world to the conditions of cognition.
Carried by Immanuel Kant; Critique of Pure Reason (1781).
Causal Powers and Dispositionalism revived Four causes
Dispositionalism deliberately revived the Aristotelian idea of real causal powers and natures, which the mechanistic and Humean traditions had rejected, arguing that properties have intrinsic causal essences.
Carried by Rom Harré; George Molnar.
Probabilistic Causation grew out of Humean Regularity Theory
Probabilistic causation generalized the regularity idea from universal conjunctions to probability-raising: Reichenbach and Suppes treated causes as events that raise the probability of their effects, extending the Humean regularity framework to indeterministic contexts.
Carried by Hans Reichenbach; Patrick Suppes.
Causal Powers and Dispositionalism reacted against Humean Regularity Theory
The revival of causal powers and dispositionalism was a reaction against the Humean regularity tradition, arguing that laws and regularities depend on real dispositional properties rather than the reverse.
Carried by Rom Harré; Roy Bhaskar; George Molnar.
Counterfactual Theory reacted against Humean Regularity Theory
Lewis presented counterfactual dependence as a superior alternative to the regularity theory, which he and others found unable to distinguish genuine causes from accidental regularities and spurious correlations.
Carried by David Lewis; Causation (1973).
Causal model grew out of Probabilistic Causation
Causal models and Bayesian networks grew out of probabilistic causation: Pearl and Spirtes, Glymour and Scheines formalized screening-off, common causes and conditional independence into graphical models for inferring causal structure from probability distributions.
Carried by Judea Pearl; Peter Spirtes; Clark Glymour.
Interventionist and manipulationist accounts grew out of Counterfactual Theory
Interventionist accounts developed the counterfactual approach by analyzing causal claims as claims about what would happen under an intervention, using possible-worlds counterfactuals to define manipulation.
Carried by James Woodward; Judea Pearl.
Interventionist and manipulationist accounts grew out of Causal model
Interventionist and manipulationist accounts were built directly on causal models: Pearl's do-calculus and Woodward's interventionist theory formalize causation in terms of interventions on variables within a causal graph.
Carried by Judea Pearl; James Woodward.
Causal Pluralism reacted against Counterfactual Theory
Causal pluralism arose partly as a reaction to the dominance of the counterfactual theory, arguing that no single analysis captures the diverse ways causation is understood across contexts.
Carried by Nancy Cartwright.
Beyond the field
Where Philosophy Of Causation's ideas came from in other fields, and what they led to there: the strongest 12 of 43 links.
Mechanistic Causation reacted against Aristotelian natural philosophy in Philosophy Of Physics
The mechanical philosophy defined itself against Aristotelian natural philosophy, rejecting substantial forms and final causes in favour of matter and motion. Bacon, Descartes and Boyle framed their causal accounts as replacements for the Aristotelian scheme.
Carried by Francis Bacon; René Descartes; Robert Boyle.
Mechanistic Causation reacted against Aristotelian physics in Classical Mechanics
Mechanistic causation arose as a rejection of Aristotelian physics: the four causes and natural-place motion were replaced by contact action and laws of motion. The new programme was articulated polemically against the Aristotelian schools.
Carried by René Descartes; Robert Boyle.
Humean Regularity Theory grew out of Empiricism in Metaphysics
Hume's regularity theory is a development of empiricist epistemology: all ideas derive from impressions, so the idea of necessary connection must be traced to a felt transition of the mind. The account of causation is an application of the empiricist principle of meaning.
Carried by David Hume; A Treatise of Human Nature (1739).
Humean skepticism about necessary connection grew out of Empiricism in Epistemology
Hume's skeptical argument about necessary connection rests on the empiricist principle that every idea must be copied from an impression. The critique of causal necessity is an application of Lockean and Berkeleyan empiricism to the idea of power.
Carried by David Hume; John Locke.
Probabilistic Causation influenced by Bayesian probability in Probability Theory
Bayes's theorem and the Bayesian account of probability supplied the formal apparatus for probabilistic theories of causation. Reichenbach and Suppes used conditional probability and screening-off, drawn from Bayesian probability theory, to analyse causal relevance.
Carried by Thomas Bayes; Hans Reichenbach; Patrick Suppes.
Counterfactual Theory influenced Counterfactual (Potential Outcomes) Framework in Epidemiological Methods
The counterfactual framework in epidemiology and statistics took its core notion of causation as difference between what would happen under alternative conditions from the philosophical counterfactual theory of causation developed by Lewis and others. Rubin and Holland explicitly built the potential-outcomes model on that counterfactual semantics.
Carried by David Lewis; Paul Holland; Donald Rubin.
Counterfactual Theory influenced by Counterfactual theory of modality in Modality
Lewis's counterfactual theory of causation drew directly on the possible-worlds semantics for counterfactuals he had developed for modality. The causal analysis applies that modal machinery to the relation between events.
Carried by David Lewis.
Causal model influenced Interventionist theory of explanation in Scientific Explanation
Causal models and the structural-equations framework, developed by Pearl and others, supplied the formal apparatus of interventions and directed graphs on which the interventionist theory of explanation rests.
Carried by Judea Pearl.
Causal model grew out of Causal graph in Causal Inference
The causal model of Pearl and Spirtes and colleagues is built on directed acyclic causal graphs and their d-separation criteria. The graph formalism is the core representational device from which the causal model framework was developed.
Carried by Judea Pearl; Peter Spirtes; Causation, Prediction, and Search (1993).
Interventionist and manipulationist accounts influenced by Potential-Outcomes and Causal Inference Framework in Philosophy Of Social Science
The potential-outcomes framework's notion of treatment assignment and manipulation supplied the core idea of intervention. Interventionist accounts formalised manipulation of a variable as the defining mark of causal relation.
Carried by Judea Pearl; James Woodward.
Four causes led to Demonstrative Science in Justification Theory
Aristotle's Posterior Analytics derives demonstrative science from the theory of causes: a demonstration is a syllogism that states the reason why, i.e. the efficient or final cause. The account of scientific knowledge is built on the four-cause framework.
Carried by Aristotle; Posterior Analytics.
Four causes led to Aristotelian Practical Reason in Philosophy Of Action
Aristotle's practical syllogism and his account of deliberation and choice apply the same four-cause apparatus (especially the final cause) to action in the Nicomachean Ethics III and VI. The practical-reason doctrine is a branch grown from the causal scheme.
Carried by Aristotle; Nicomachean Ethics.

Philosophy Of Causation is a planet in Nousfarer
Its city has a tower for each of these 12 frameworks. Land, walk in, and climb one concept at a time.
Land on Philosophy Of CausationRelated fields
Fields its ideas are linked to
- Mental Causation And Physicalism 6 links
- Free Will 4 links
- Scientific Explanation 4 links
- Philosophy Of Science 3 links
- Epidemiological Methods 2 links
- Mathematical Explanation 2 links
- Philosophy Of Action 2 links
- Religion And Science 2 links
- Physics 2 links
- Christian Philosophy 1 link
- Data Science 1 link
- Probability Theory 1 link
More in Metaphysics
- Free Will
- Grounding And Fundamentality
- Metaphysics
- Modality
- Ontology
- Persistence And Material Objects
- Personal Identity
- Philosophy Of Time
- Properties And Universals
- Social Ontology
See where Philosophy Of Causation sits on the Map of Human Knowledge