# Intervention and experiment

## TL;DR

This paper challenges the idea that scientific experiments must involve active intervention, showing that non-interventionist studies can also be proper experiments.

## Contribution

The paper argues that non-interventionist studies can be classified as proper experiments, challenging the traditional view of experimentation.

## Key findings

- Non-interventionist studies can be epistemically as valuable as interventionist experiments.
- Intervention is not necessary for uncovering causal structures or producing strong evidence.
- Some non-interventionist studies should be considered proper experiments.

## Abstract

The received view of scientific experimentation holds that science is characterized by experiment and experiment is characterized by active intervention on the system of interest. Although versions of this view are widely held, they have seldom been explicitly defended. The present essay reconstructs and defuses two arguments in defense of the received view: first, that intervention is necessary for uncovering causal structures, and second, that intervention conduces to better evidence. By examining a range of non-interventionist studies from across the sciences, I conclude that interventionist experiments are not, ceteris paribus, epistemically superior to non-interventionist studies and that the latter may thus be classified as experiment proper. My analysis explains why intervention remains valuable while at the same time elevating the status of some non-interventionist studies to that of experiment proper.