Cognitive Ghosts

Episode Summary

Executive Summary: The episode explores “cognitive ghosts”: strange experiences caused by mismatches in memory, perception, and bodily awareness, including déjà vu, presque-vu, jamais vu, blindsight, the call of the void, cute aggression, hypnic jerks, and end-of-life visions. It also widens the idea to ghostly cultural traces—tropes and myths that outlive their origins and shape what we think is real.

Main Topics: Déjà vu and laboratory analogues (Priority: 5/5): The hosts use a word-list trick to induce a false feeling of familiarity, explaining how this lab setup can approximate déjà vu and show that familiarity and memory are separable processes. Presque-vu and jamais vu (Priority: 5/5): They discuss tip-of-the-tongue states and the uncanny loss of meaning from repeated words, framing both as cognitive failures where the brain’s helping systems overshoot or misfire. Blindsight and embodied ghostly behavior (Priority: 5/5): A discussion of how brain damage can leave people visually responsive without conscious awareness, revealing hidden processing systems that act before the self can explain them. High place phenomenon, cute aggression, and hypnic jerk (Priority: 4/5): These everyday eerie sensations are presented as defensive or regulatory responses—fear-generated impulses, overarousal corrections, and sleep-transition misfires. Cultural cognitive ghosts and media tropes (Priority: 4/5): The episode argues that films, TV, and inherited stories create fake but sticky ‘realities’—like electrocution visuals, spy earpiece touches, or salt/amber myths—that persist because they are useful or familiar. Ghostly companions and end-of-life visions (Priority: 5/5): Hannah introduces literal-seeming ‘ghosts’ from extreme environments and dying patients, suggesting they may arise from sensory deprivation, brain-body mismatch, or end-of-life neural changes.

Key Arguments: Déjà vu may be your brain’s fact-checking system detecting familiarity without being able to locate the source of the memory. Presque-vu shows that knowing something is in memory and being able to retrieve it are different processes. Jamais vu may function as a protective mechanism that breaks repetitive neural loops by making the familiar feel strange. Blindsight demonstrates that visual information can guide behavior without entering conscious awareness, proving the self is not the whole controller. High place phenomenon and cute aggression can be understood as the brain generating counter-signals or phantom explanations to rebalance overwhelming states. Cultural tropes become ‘ghosts of reality’ because representations in media can be more common than the underlying real-world experience. The ‘third man’ phenomenon and similar presence illusions may emerge from disrupted body schema and proprioception under stress, fatigue, or oxygen deprivation. End-of-life dreams and visions may reflect structured brain activity near death rather than random delirium, though their exact cause remains unknown.

Data Points: Human genome typing scale: 60 words per minute for about 50 years - Used to illustrate the scale of DNA instructions in each cell Cancer types: More than 200 - Cancer is described as a collection of more than 200 distinct diseases False S-word count in the list experiment: 0 - Michael correctly identifies that none of the words in the initial list begin with S Laboratory-induced jamais vu repetition: About 30 repetitions - Repeating a word around 30 times can trigger feelings of strangeness Jamais vu study response rate: About 70% - In the laboratory repetition study, around 70% stopped because they felt peculiar Typical baseline for inducing jamais vu: 30 repetitions - Mentioned as enough to make a normal word feel weird Delay in body-ownership experiment: 500 milliseconds - A half-second delay in the robotic back-poking setup was enough to create a sense of another presence Presence-illusion participants: Loads of participants / multiple ghosts reported - Healthy volunteers in the delayed-feedback lab experiment reported feeling someone behind them London tube call-of-the-void prevalence: 60% - Michael cites a figure that 60% of Londoners report the impulse-feeling on the tube Cute-aggression coding study: Responses were coded from participant speech - Researchers showed people cute stimuli and categorized utterances as caretaking, appraisal, approach, overwhelm, and aggression End-of-life dreams and visions prevalence: Up to 88% - Christopher Kerr’s work with dying patients found most experienced at least one vivid dream/vision End-of-life sample size: 1,400 patients - Kerr’s hospice and palliative-care dataset over a decade Duration of end-of-life data collection: A decade - Describes the time span over which the patient stories were collected Ancient wine dilution ratio: 1 to 2 - Used in the amethyst goblet anecdote to explain the origin of a modern myth Historical support for soldier pay and salt: Solarium / salary origin - Explains that salary comes from Roman payments intended to buy salt Brain-wave pattern near death: Gamma-wave surge - Reported in EEG observations of dying patients, associated with intense concentration and memory retrieval

Pivotal Quotes: "It’s your brain glitching, essentially, saying, wait, no, have we been here before?" — Hannah Fry: Explaining the leading interpretation of déjà vu as a familiarity/memory mismatch "Maybe we’re just like the sum of ghosts rather than like a single individual with all of these things spinning off it." — Michael Stevens: Reflecting on whether there is a single central self controlling all cognition "there are these ghosts inside that are doing a bunch of stuff without your authority and control." — Hannah Fry: Summarizing the episode’s core idea that unconscious systems shape behavior

Implications: Listeners are encouraged to see strange mental experiences as meaningful brain processes, not just oddities. For science and medicine, the episode highlights why studying edge cases matters and why some ‘ghosts’ may reveal how perception, memory, and identity really work.