Lesion Studies: Learning from Brain Damage

How studying people with brain injuries — Leborgne, Gage, H.M. and thousands since — revealed what brain regions do, the logic of double dissociation, modern lesion-mapping, and the method's pitfalls.

Long before brain scanners, the main way to learn what a brain region did was to observe what was lost when it was damaged.

Classic cases

Broca's patient Leborgne ("Tan") linked the left frontal lobe to speech production (1861). Phineas Gage linked the frontal lobes to personality and self-control. Henry Molaison revealed the role of the medial temporal lobes in memory. War injuries added enormously: after the Russo-Japanese War and the First World War, studies of soldiers with bullet wounds mapped the visual field onto the occipital cortex in detail.

The logic of dissociation

If damage to region A impairs ability X but not Y, that's a single dissociation — suggestive but not conclusive (X might just be harder). A double dissociation is stronger: damage to A impairs X but not Y, while damage to B impairs Y but not X. Broca's and Wernicke's aphasias form a classic double dissociation between producing and understanding language, as do impairments of recognising faces versus objects.

Modern lesion mapping

Today, voxel-based lesion-symptom mapping uses MRI scans from many patients to test, across the whole brain, where damage is statistically associated with a particular symptom. Lesion network mapping, developed by Michael Fox and colleagues, maps lesions onto a reference connectome to find networks — rather than single spots — linked to symptoms such as delusions or addiction remission. Patients who have had surgery for epilepsy or tumours provide carefully documented lesions.

Experimental lesions and virtual lesions

In animals, researchers make precise lesions or, increasingly, use reversible methods: cooling, drugs, optogenetics, chemogenetics. In humans, TMS can create brief, harmless "virtual lesions".

Pitfalls

Natural lesions rarely respect functional boundaries — a stroke follows blood vessels, not brain areas. The brain reorganises after injury, so effects change over time. Damage can disrupt fibres passing through a region (affecting distant areas) and cause diaschisis: reduced function in distant connected regions. And a region being necessary doesn't mean it performs the function alone. Despite these challenges, lesion evidence remains uniquely powerful for showing what the brain needs.