Memory systems: facts, episodes and skills
Distinguish episodic, semantic and procedural performance and interpret a selective impairment without locating all memory in one structure.
Question and prediction
If someone cannot describe a practice session, can their performance still improve?
Predict whether explicit recall and practiced skill must rise or fall together.
Mechanism and evidence
Episodic memory concerns events situated in one’s experience; semantic memory concerns knowledge such as meanings and facts. Both are commonly grouped as declarative or explicit memory. Procedural learning concerns practiced skills and habits, while other nondeclarative effects include priming and conditioning. These classifications organize evidence; they do not mean each task uses only one system or that memories are literal files in separate drawers.
Scoville and Milner’s 1957 clinical case series described severe recent-memory difficulties after extensive bilateral medial temporal surgery. This authentic evidence helped challenge a view of memory as one undifferentiated ability. The operations affected multiple structures, and clinical cases were not randomized lesion experiments. Do not infer that the hippocampus stores every memory or cite the 1957 report as if it contained all later findings about skill learning.
The fictional task battery separates what is measured. Across practice, tracing errors decrease while recall of the preceding session remains low. That pattern is compatible with different contributions to skill learning and event recollection. It does not establish a diagnosis, a uniquely localized brain mechanism or the absence of all explicit knowledge. Motor ability, instructions and scoring must also be considered.
A dissociation is a difference in how tasks are affected. One task may simply be easier, so stronger evidence compares difficulty and multiple measures; a double dissociation asks whether a different pattern appears under another condition. The unfamiliar task asks for a discriminating observation, not a medical assessment. Use invented records and the historical report without attempting to test anyone’s memory health.
Inspect the relationships
| Task | Outcome | Interpretation limit |
|---|
| Tracing | Errors decrease | May combine motor practice and strategy |
| Previous-session questions | 1/5 correct | Wording and retrieval demands matter |
| Fact knowledge | Not measured | Cannot infer preservation or loss |
Worked example
Tracing errors decline by 11 from first to third session, while event-question accuracy remains 20%. This is a difference between task outcomes, not proof that one biological structure alone caused it. Changing question wording creates an additional task-demand explanation.
Your investigation
Classify three outcomes—event recollection, factual knowledge and practiced skill—by what each task measures. Use the historical case card to distinguish clinical evidence from the fictional battery. Write 200 words proposing a second task that would test the difficulty explanation.
Unfamiliar transfer
One task is much harder than another. Why does this matter for a dissociation? Explain your reasoning before checking feedback.
Reading
Scoville and Milner (1957): loss of recent memory after bilateral hippocampal lesions. The records below are original fictional teaching cases, not research results from this source.
Fictional task battery Across three sessions, tracing errors are 18, 12 and 7. Correct answers about the previous session remain 1 of 5. Materials and scoring are held constant in this teaching case.
Fictional alternative The tracing task becomes familiar, while the session questions change wording each time. No brain data or clinical history is supplied.
Authentic evidence and its limits
Clinical case-series evidence — William B. Scoville and Brenda Milner (1957). The original report described marked recent-memory difficulties after extensive bilateral medial temporal resections and related their extent to the affected structures.
Bounded paraphrase of a clinical case series, not randomized damage or a claim that all memory resides in one site. Later skill-learning findings should not be silently attributed to this report. Historical terminology should be read in context.
Read the original source. Original LearnAI paraphrase and attributed factual results; no reproduced figure or substantial verbatim passage. Original publication retains its rights.
