You cannot open a healthy brain to see where memory lives. But when one part of a brain stops working and one ability disappears with it, you learn something no laboratory task could tell you. The case of HM is the clearest example psychology has.
📘 What you need to know
The brain has two hemispheres, covered by a thin layer called the cerebral cortex.
There are four main lobes — frontal, parietal, occipital and temporal — each divided between the hemispheres, so each lobe mirrors itself on the other side.
Cognitive neuroscience is the scientific study of brain structure and function and how these relate to behaviour.
It is central to localisation of brain function: mapping which structures link to which behaviours.
If a structure is damaged, the functions associated with it may become impaired.
Milner (1958); Scoville (1997) studied HM, who had two-thirds of his hippocampus removed.
HM had severe anterograde amnesia and partial retrograde amnesia.
Where things are, roughly
Deliberately schematic — real lobes are not neat blocks. What matters for the exam is that the hippocampus sits deep within the temporal lobe, which is exactly what HM’s surgery removed.
Localisation is easiest to see through damage. Damage to Broca’s area in the left hemisphere leaves someone with difficulty producing fluent speech. Damage to the pre-frontal cortex can change personality, sometimes making a person more aggressive or impulsive.
This is why case studies of brain damage matter so much. Nobody could ethically remove part of a healthy hippocampus to see what happens. When surgery or injury does it anyway, psychology gets an answer it could not otherwise have.
Key study: Milner (1958); Scoville (1997)
🧩 The study in four steps
Aim — to investigate memory loss in a brain-damaged patient known as HM, through examination of his brain post-mortem and via MRI.
Participant — HM (Henry Molaison) was run over by a bicycle aged nine and developed epileptic seizures. At 27 he had a bilateral medial temporal lobe resection, removing about two-thirds of his hippocampus. His epilepsy improved, but he was left with extreme anterograde amnesia and partial retrograde amnesia.
Procedure — Brenda Milner visited frequently, administering personality and mood tests, depression questionnaires and interviews with psychiatrists. His scores showed no depression, anxiety or psychosis, and he was aware of his own condition. His IQ was normal, but the Wechsler Memory Scale showed severe memory impairment. MRI scans in 1992–1993 confirmed the medial temporal lobe had been particularly affected. Milner and Corkin studied him for over 50 years until his death at 82.
Findings and conclusion — memory is not simply part of a bundle of cognitive functions sitting in the cortex. It is a distinct function, localised to the temporal lobe and specifically the hippocampus. Hippocampal damage may be linked to long-term anterograde amnesia, and the formation of new memories is linked to the hippocampus.
The right-hand column is the part students forget to mention. If HM had lost everything, the study would prove nothing about localisation — it is the contrast that carries the argument.
Two words to keep apart.Anterograde amnesia is losing the ability to form new memories going forward. Retrograde amnesia is losing memories from before the damage. HM had extreme anterograde and only partial retrograde amnesia.
Evaluating the study
Strengths
Mixed methods. The case study used both qualitative and quantitative methods, generating data that was both reliable and explanatory.
Groundbreaking. It gave enormous insight into localised regions of the brain and reshaped how memory was understood.
Longitudinal depth. Over 50 years of study allows a level of detail no single session could produce.
Limitations
Confounding variable. HM’s brain may already have been damaged by his epilepsy, which lowers the validity of attributing everything to the surgery.
Limits of knowledge. Not enough is known about the brain to draw iron-clad conclusions about exactly which functions are affected by which damage.
One participant. A single case cannot be generalised confidently to everyone.
Link to concepts
Bias — case studies involve researchers working closely with one participant over long periods, which risks researcher bias affecting objectivity and impartiality. Getting too close can lead to over-reporting some findings and under-reporting others, perhaps out of a sense of duty towards the participant. Researchers must practise reflexivity to keep the results credible.
Responsibility — researchers must be alert to the potential for harm to a vulnerable person. HM had no recollection of recent events, did not even recognise the researchers, and lived under supervision. Informed consent is the hardest problem here: how could HM consent to something he would instantly forget? His mother initially gave consent by proxy, followed by a court-appointed conservator after her death. His family later contested this, arguing he never willingly consented — which shows how sensitive this kind of research is.
Exam practice
WORKED EXAMPLE
Explain one study of the effect of brain damage on memory. [3 marks]
1. Participant and damage
HM had about two-thirds of his hippocampus removed in a bilateral medial temporal lobe resection at 27.
2. What was measuredNormal IQ, severely impaired Wechsler Memory Scale scores, alongside interviews and MRI.
3. What it shows
Memory is a distinct function localised to the hippocampus, not spread evenly through the cortex.
Damage + measure + conclusionThe normal IQ is the detail that makes the conclusion work. Never leave it out.
WORKED EXAMPLE
Discuss the use of case studies in cognitive psychology. [ERQ paragraph]
Strength
The HM case gave depth impossible in a lab, combining qualitative interviews with quantitative test scores over 50 years.
LimitationHowever one participant cannot be generalised from, and his epilepsy is a confounding variable.
Ethical dimension
Consent by proxy was used because HM could not retain the information needed to consent himself, and his family later contested it.
Judgement
Case studies of rare damage buy irreplaceable insight, but at a real cost in generalisability and ethical certainty.
Irreplaceable, but not repeatableThe consent problem is specific to HM. Generic ethics points score much less.
💡 Exam tip
Name the surgery properly: bilateral medial temporal lobe resection. It is a mouthful worth learning.
Give one thing HM lost and one thing he kept. The contrast is the argument.
Say localisation of function explicitly — it is what the study is actually evidence for.
Use HM again for the multi-store model, where he supports separate stores for STM and LTM.
Mention the MRI in 1992–1993 if the question is about methods — it shows triangulation across decades.
Consent by proxy is a strong, specific ethics term. Use it rather than “it was unethical”.
⚠️ Common mix-up
Anterograde and retrograde reversed. Anterograde is forwards, from the damage onwards.
HM was not unintelligent. His IQ was normal throughout.
The hippocampus is not “where memories are stored”. It is needed to form new long-term memories.
The whole hippocampus was not removed. About two-thirds of it was.
This is not an experiment. It is a longitudinal case study.
Use his initials or full name respectfully. He was a person, not a specimen.
Up next: Atkinson and Shiffrin’s Multi-Store Model — if memory really is a distinct system, what shape does that system have? The first serious attempt at an answer came in 1968.
Want this explained one-to-one?
Book a free session with an experienced IB Psychology tutor and get your trickiest topics made simple.
