---
title: Dispositionalism (Tendencies, Powers)
type: concept
tags: [philosophy-of-science, causation, dispositionalism, powers, tendencies, anjum]
sources: ["[[sources/2023-anjum-rocca-phi403-causation-in-science]]", "[[sources/2023-anjum-rocca-phi403-lecture-09-from-regularities-to-tendencies]]"]
created: 2026-04-20
updated: 2026-04-20
---

# Dispositionalism

The view — defended in the course by [[entities/rani-lill-anjum]] — that **properties have real causal powers or dispositions**, and causation happens when dispositions *manifest*. Roots in Aristotle and Aquinas; revived by Mumford, Anjum, Cartwright, C.B. Martin.

## Core commitments
- **The dispositional modality** — a third modality between pure contingency (Hume) and necessity: causes **tend or dispose** toward their effects ([[sources/2023-anjum-rocca-phi403-lecture-09-from-regularities-to-tendencies]]).
- **Dispositions can exist unmanifested** — dynamite is explosive whether or not it explodes; an apple seed's cyanide is poisonous even if it never poisons anyone.
- **Tendencies come in degrees**, can be **counteracted**, **enforced**, or overridden by other dispositions.
- **Mutual manifestation partners** (C.B. Martin) — a fire needs flammable material *and* oxygen; none of the partners could produce the effect alone ([[sources/2023-anjum-rocca-phi403-lecture-07-one-effect-one-cause]]).
- **Singularism** — each causal set-up is unique; causation can occur in a one-off instance (Big Bang), so repetition is not required for causation ([[sources/2023-anjum-rocca-phi403-lecture-11-is-more-data-better]]).

## Epistemological implications
- Tendencies are **not reducible to statistical frequencies** for several reasons: [[concepts/rct-limitations|ecological fallacy]] (individual ≠ average), individual variation, unique events, strong-but-rare tendencies, weak-but-frequent tendencies.
- Therefore, qualitative methods, case studies, mechanistic reasoning and N=1 studies can be *more* evidential than large-N statistics for some causal questions.
- Predictions become **tendency claims** — "*C disposes toward E*" — which are fallible but informative, not vacuously true ([[sources/2023-anjum-rocca-phi403-lecture-18-risky-predictions]]).

## Rivals
- [[concepts/regularity-theory-of-causation]] (Hume) — the main rival on the universalist side.
- **Conditional necessity** — a weaker tendency-talk on which causes *necessitate* effects when all conditions are right. Dispositionalism denies this: even in ideal conditions, manifestation is not guaranteed.

## Relevance to BPM
Tendency-framed predictions fit process monitoring well: a PPM model outputting *"case X tends toward SLA violation given its current prefix"* is a dispositional claim, not a regularity or counterfactual claim. This is explicit in how tendencies are framed in [[sources/2023-anjum-rocca-phi403-lecture-18-risky-predictions]].

## Related
[[concepts/causation]] · [[concepts/regularity-theory-of-causation]] · [[concepts/probabilistic-causation]] (propensity theory is dispositionalism's probabilistic cousin) · [[concepts/mechanisms-in-science]] · [[concepts/rct-limitations]]