---
title: "PHI403 Lecture 09 — From Regularities to Tendencies"
type: source
tags: [philosophy-of-science, dispositionalism, modality, powers, tendencies, aristotle]
authors: [Anjum, Rani Lill; Rocca, Elena]
year: 2023
venue: "PHI403 Causation in Science, NMBU"
kind: handout
raw_path: "raw/Philosophy of Science/PHI302 09 From Regularities to Tendencies.pdf"
created: 2026-04-20
updated: 2026-04-20
key_claims:
  - The dispositional modality is a third option between Humean pure contingency and necessity — causes tend toward their effects.
  - Dispositions can exist unmanifested (dynamite is explosive without ever exploding); they come in degrees and can be counteracted.
  - Tendencies are not reducible to statistical frequencies: ecological fallacy, individual variation, unique events, strong-but-rare and weak-but-frequent tendencies all block the reduction.
  - A "deeply tendential" view holds that even in ideal conditions the cause still only tends toward its effect — it could still fail.
---

# PHI403 Lecture 09 — From Regularities to Tendencies

Presents the course's preferred alternative to Humean regularity: **[[concepts/dispositionalism|dispositionalism]]**. The three modalities:

1. **Pure contingency** (Hume) — anything could in principle follow anything.
2. **Necessity** (many anti-Humeans) — causes necessitate effects under ideal conditions.
3. **Dispositional modality** — causes *tend* toward their effects; more than contingency, less than necessity.

Dispositions (fragility, elasticity, flammability, solubility) are real properties of things; causation occurs when they manifest with their partners. Tendencies come in degrees and can be counteracted, enforced or overridden.

The lecture then distinguishes **conditional necessity** (weak tendency talk — causes necessitate effects when conditions are right) from the **deeply tendential** view — even with all conditions met, the cause *still* only tends toward its effect. The deep view treats tendency as *internal* to the cause, not an external context-sensitive modifier.

**Crucial epistemological point**: tendencies ≠ statistical frequencies. Five reasons:
i. Correlation is not causation.
ii. Individual tendencies can vary within a population average.
iii. A tendency can be unique (Hume's example: the universe's origin).
iv. Strong individual tendencies can be statistically rare (1 in 100 000).
v. Weak tendencies can be statistically frequent yet never manifest (cyanide in apple seeds).

So statistics alone cannot reveal dispositions; mechanistic/qualitative methods are needed.

## Connections
Back-link: [[sources/2023-anjum-rocca-phi403-causation-in-science]]. Concepts: [[concepts/dispositionalism]] · [[concepts/regularity-theory-of-causation]] · [[concepts/probabilistic-causation]] · [[concepts/rct-limitations]].