Darker cells mean a higher score for this topic on that metric.
LessMore
Last reviewed Sources & creditsMedia creditsMethodology
Quick answers
Does this atlas rank groups by IQ, race or nation?
No. Pathways and evidence stay individual. Scores are browsing aids — never group biological or racial ranks.
Is every autistic person a savant?
No. Most are not. Most savants in the modern clinic have autism or other developmental/brain difference — the overlap is not identity.
Can savant skills be trained like Ericsson hours?
The island is usually not a conservatoire story. Trainability is low; support and harm-reduction are the educational tasks. Some skills refine with opportunity; they do not become general talent.
Is this giftedness?
No. Gagné/Renzulli school giftedness is a different file. Jaggedness can appear in both; the clinical island-plus-disability pattern is specific.
What about Rain Man?
A film that mixed Peek-like memory with fiction. Use it as a media-warning, not a case file.
Did Gardner use savants to prove MI?
Yes, as existence proofs of unevenness. That does not validate eight school IQs (Waterhouse).
Research on savants runs from Darold Treffert (1960s–2010) through later critiques. Mechanisms are interacting systems.
Evidence percentages below are relative judgements for this atlas, not fabricated precise statistics. Named studies keep their published bands.
Findings
- 1960s–2010 · Darold Treffert — A clinical pattern with islands Prevalence is low; music, art, calendar, calculation, memory dominate the islands.
- 1887 · J. Langdon Down — Early clinical notice In the language of his century — history, not our vocabulary for people.
- 1988 · Media studies of Rain Man — A template that overgeneralises Most autistic people are not Peek; most savants are not the film.
- 1983 · Howard Gardner — Savants as unevenness proofs A philosophical use; not eight IQs (see MI page).
- 2000s · Neuroimaging case studies — Atypical networks in some individuals Small n; not a racial or group story; not a party trick map.
- 2010s · Music-teaching reports (Paravicini-type) — Support can refine an island Refinement ≠ turning the life into a general gifted track.
Mechanisms
- Narrow, atypical specialisation — An island, not a timetable of school subjects.
- Different memory/perception routes — Sometimes described in cases (Peek’s memory; some graphic memory) — do not generalise to a superpower brand.
- Low far transfer — The defining educational fact: the island does not become “talent at life”.
- Environment as harm or help — Noise, consent, sleep, exploitation — the real levers, not 10,000-hour posters.
Evidence
Best clinical description we have; n still small.
Rain Man as folklore.
Wrong model for most islands.
Unevenness ≠ eight tests.
Unethical and unscientific.
Debates
- Word choice — Savant versus abandoned slurs versus “exceptional skill”. This page keeps Treffert’s clinical word and his dignity rule.
- Public performance — Wiltshire and Lemke files: art versus extraction. Consent is the debate.
- How the island relates to autism — Overlap is real and partial. Do not collapse the identities.
Savants: size the claim to the study. Interaction, not a 50/50 slogan, is the default model.
OriginsScienceSignsDevelopPracticeExemplarsMyths In this language region