Chapter 4 turns to the so-called general semantic network (GSN), which consists of several very high-level, interconnected cortical areas that are considered transmodal because they operate on two or more types of sensory/motor information. Mounting evidence suggests that the GSN contributes to word meanings by performing a variety of integrative and relational functions. Continuing with the example of scissors, the GSN conjoins its distinctive experiential features (which, as noted above, are housed in modal networks of sensory/motor areas), assembles them into a coherent concept that can easily be accessed during language production and comprehension, situates this particular concept in relation to many others, and connects it with personal memories and attitudes (e.g., using one’s favorite pair of scissors in an enjoyable project). As we’ll see, one of the most fascinating properties of the GSN is that it overlaps substantially with the well-studied default mode network (DMN), which mediates internally oriented mental processes such as recalling one’s past, imagining one’s future, simulating dialogues, and contemplating other people’s thoughts and feelings – all of which require semantic cognition.

Sleep paralysis is one of the most terrifying experiences a person can have—and it’s surprisingly common. Cultures around the world describe eerily similar episodes: waking up unable to move, a crushing pressure on the chest, and the overwhelming sense that someone—or something—is in the room. This chapter explores how those experiences may arise from the collision of sleep architecture and perceptual ambiguity. It covers the basic neurobiology of REM sleep, explains what happens when paralysis persists into wakefulness, and investigates how hallucinations can emerge in these liminal states. The chapter also examines the role of the temporoparietal junction in out-of-body experiences and the sensation of a nearby presence. Rooted in both science and cultural context, this chapter offers a grounded explanation for a deeply human phenomenon—one that’s haunted people for centuries and continues to blur the line between brain and belief.

Chapter 5 addresses yet another aspect of word meanings. Back in the mid-twentieth century, the linguist J. R. Firth (1957, p. 11) stated that “you shall know a word by the company it keeps.” More recently, this idea has been supported by distributional semantic models (DSMs), which come from computational linguistics and demonstrate that a word’s meaning can in fact be derived partly from its statistical co-occurrence patterns with other words. For instance, part of the meaning of scissors can be derived from its tendency to be used together with certain other words like sharp, pointy, cut, snip, paper, hair, etc. DSMs are surprisingly good at predicting people’s performance on many (although not all) conceptual tasks, and they are now so sophisticated that they constitute the engines of many chatbots and AI systems. What’s more, by combining DSMs with brain mapping methods, a rapidly growing line of research has been accumulating evidence that the distributionally based properties of word meanings are not only captured by purely verbal representations in the core language network, but enable a “quick and dirty” shortcut to comprehension.

If we’re trying to come up with a theory to explain the sound of footsteps behind you, a feeling of a presence, lights that you can’t explain, or the psychic who knows everything about you, we might be tempted to say that supernatural forces are at work. But we also know that each one of these instances can be easily explained with neuroscience and psychology. This is what I’ve attempted to do in this book.

The Science of the Supernatural might, at first, feel like an oxymoron. I don’t think most people would immediately see the myriad connections between the paranormal and psychology. I didn’t at first, either. I’ve always loved ghost stories, horror movies, and scary novels. I have a distinct memory of lying in my bed as a kid, trying unsuccessfully to go to sleep. I had just read Stephen King’s short story “The Boogeyman.” I remember staring at my closet door, sure that it was slowly creaking open. Certain that the boogeyman was on the other side, waiting to kill me.

Chapter 1 sets the stage by describing several linguistic and psychological aspects of word meaning, with emphasis on those that have received the most attention in cognitive neuroscience. Specific topics include the treatment of word meanings as public concepts for social coordination; the decomposition of word meanings into semantic features; the characterization of word meanings in terms of frames, prototypes, mental models, and background situations; the nature of word associations and co-occurrence patterns; the influence of context on interpretation; and the importance of crosslinguistic similarities and differences.

Zombie myths have captured imaginations for centuries, but their roots may lie in real-world infections that alter behavior in terrifying ways. This chapter explores the biological underpinnings of the zombie archetype, beginning with cultural practices surrounding Haitian Vodou and moving into the realm of neuroscience and virology. Rabies serves as a chilling real-world analog to zombification, with symptoms like aggression, hydrophobia, and loss of cognitive control emerging as the virus travels from the bite site to the central nervous system. The chapter also examines Toxoplasma gondii, a parasite capable of rewiring host behavior and reducing fear responses, particularly in rodents. By tracing the ways infectious agents can alter motivation, movement, and fear, this chapter offers a grounded, scientific perspective on one of the most enduring horror tropes—and explores what happens when the threat isn’t supernatural, but biological, and it’s already inside the body.

Chapter 7 concentrates on abstract words like democracy, luxury, and chance. These words are harder to characterize and investigate than concrete ones like bird, mug, and banana, but the pace of progress in understanding their cognitive and neural bases has dramatically increased in recent years. For instance, it’s now clear that compared to concrete words, abstract ones rely more on verbal associations, occur in a broader range of contexts, and are rated higher for certain types of semantic features (e.g., time, social interaction, emotion, and drive). Consistent with these differences, it’s well-established that abstract words rely more than concrete ones on a few very high-level cortical areas that play vital roles in language processing while also contributing to the GSN/DMN. And yet there’s also mounting evidence that, like concrete words, many abstract ones are anchored to some extent in systems for perception and action. In addition, an increasing amount of research has been exploring how different categories of abstract words (e.g., those for numbers, emotions, mental states, and moral judgments) are associated with different sets of partly shared and partly segregated brain regions.

Chapter 3 begins to elaborate a central theme of the book, which is that word meanings are not localized in just one part of the brain; instead, they have a widely distributed web-like layout that includes many different cortical areas and corresponding types of representation. This particular chapter focuses on the experiential (e.g., visual, auditory, and motor) features of word meanings. The key idea is that, in keeping with theories of grounded/embodied cognition, these concrete features are identical to some of the modality-specific representations that allow us to make sense of our nonlinguistic experiences involving the pertinent types of entities and events. For example, the word “scissors” denotes a kind of household tool with specifications for shape, motion, sound, and manipulation, and considerable research suggests that we store these features directly within some of the same cortical areas that are engaged when we see, hear, and use scissors. Such findings are exciting because they support the intuitive view that words are like instructions for neurally simulating experiences, albeit usually in an automatic, implicit manner. There’s still a great deal of debate, however, about the precise ways in which word meanings relate to perception and action.

Chapter 1 establishes the foundational concepts of neuroimaging by exploring the complex relationship between brain structure and mental function. It traces the historical progression from ancient surgical approaches to modern noninvasive techniques, contextualizing how technological innovations have transformed our understanding of neural processes. The chapter examines the multiscale nature of brain investigation, from single-neuron recordings to population-level measurements, and evaluates the critical tradeoffs between spatial and temporal resolution across imaging modalities. Key neurophysiological principles underlying these technologies are introduced, including neuronal action potentials, hemodynamic responses, and the chemical processes that support neural activity. The text challenges common neuromyths while addressing fundamental questions about functional organization, from modular specialization to distributed network processing. By comparing the relative strengths and limitations of major neuroimaging tools (fMRI, EEG, MEG, PET, and TMS), the chapter provides an analytical framework for understanding how these methodologies collectively advance our ability to correlate brain activity with cognitive and behavioral processes, setting the stage for more detailed exploration in subsequent chapters.

This chapter tackles positron emission tomography (PET), a functional neuroimaging technique that revolutionized brain imaging in the 1970s by providing the first colorful maps of brain activity. Beginning with its historical development from Hans Berger’s early hemodynamic measurements to modern scanners, the chapter examines how PET visualizes metabolic processes by tracking radioactively labeled tracers in the bloodstream. Unlike structural imaging methods, PET detects gamma rays emitted when positrons from the radiotracer collide with electrons, allowing researchers to measure regional changes in blood flow, glucose metabolism, and neurotransmitter activity related to cognitive processes. The chapter details practical aspects of PET studies, including experimental design, data acquisition, image reconstruction techniques, and visualization methods like subtraction analysis for mapping task-related brain activity. While MRI-based techniques have supplanted PET for many cognitive neuroscience applications, PET remains invaluable for certain investigations due to its unique ability to label diverse compounds, particularly for studying neuropsychiatric disorders, neurotransmitter systems, and metabolic processes in diseases like Alzheimer’s and epilepsy.

Chapter 7 deals with neuroimaging methods for investigating the structural components underlying brain function. Beginning with lesion-symptom mapping (LSM), which identifies relationships between localized brain damage and specific cognitive deficits, the chapter examines how structural abnormalities correlate with functional impairments. Three primary approaches to measuring brain structures with MRI are discussed: structure tracing for hypothesis-driven volumetric analysis, voxel-based morphometry (VBM) for whole-brain comparison of tissue concentration, and surface-based morphometry (SBM) for analyzing the cortical sheet’s unique properties including thickness, curvature, and gyrification. The chapter then explores diffusion tensor imaging (DTI), a technique that visualizes white-matter tracts by measuring the anisotropic diffusion of water molecules along axon bundles. DTI tractography reveals the brain’s “highways,” short, intermediate, and long-range fiber pathways that connect functional modules within and across hemispheres. Together, these complementary techniques provide critical insights into the structural architecture supporting brain networks, offering a more complete understanding of brain organization when combined with functional imaging methods.

Chapter 9 introduces transcranial magnetic stimulation (TMS), a neurostimulation technique that uses rapidly changing magnetic fields to induce electric currents in targeted brain regions. Beginning with its historical roots in 19th-century electromagnetic experiments and evolving through Anthony Barker’s groundbreaking 1985 demonstration, TMS has become a critical tool for establishing causal relationships between brain activity and behavior. Unlike neuroimaging methods that only observe brain activity, TMS can temporarily interrupt or enhance neural processing, enabling researchers to create “virtual lesions” and directly test hypotheses about regional brain function. The chapter examines TMS delivery methods, single-pulse, paired-pulse, and repetitive stimulation, and their differential effects on cortical excitability. It details four primary research applications: virtual lesions for establishing causality, chronometry for determining processing timelines, mapping functional connectivity between brain regions, and tracking neuroplasticity. Clinical applications are discussed, particularly for treating depression and presurgical mapping. The chapter also addresses practical aspects of TMS implementation, localization techniques, and safety considerations, concluding with a brief overview of transcranial direct current stimulation (tDCS) as a milder alternative stimulation approach.

Chapter 2 traces the development of electroencephalography (EEG) from its inception with Richard Caton’s pioneering work in 1875 to its current status as a cornerstone of human neuroimaging. The chapter discusses how EEG captures the electrical signals generated by synchronous activity of pyramidal neurons arranged in open fields perpendicular to the cortical surface. It examines the technical evolution of recording systems, from basic silver-chloride electrodes to modern active electrode arrays with built-in amplification, and explains the standardized 10-20 electrode placement system that enables spatial mapping of brain activity. The chapter addresses the inverse problem that constrains EEG’s spatial resolution while highlighting its exceptional temporal precision for tracking neuronal events in millisecond timescales. Special attention is given to the characteristic oscillatory patterns in different frequency bands (alpha, beta, theta, delta, gamma) and their association with cognitive states ranging from deep sleep to focused attention. The chapter details practical considerations for obtaining clean recordings, including artifact reduction techniques and experimental design. By evaluating EEG’s strengths (temporal precision, direct measurement of neural activity, accessibility) alongside its limitations, the chapter positions EEG as an enduring, versatile tool for both clinical applications and cognitive neuroscience research despite technological advances in other imaging modalities.

This chapter examines intracranial electroencephalography (iEEG), a rare but powerful technique offering unparalleled insights into human brain function by recording electrical activity directly from the brain’s surface. It traces iEEG’s development from pioneering work by Penfield and Jasper in the 1950s to modern applications with up to 1,024 recording channels. The chapter outlines the two primary surgical approaches, stereo EEG with depth electrodes and electrocorticography with surface grids, and explains how these techniques achieve both high temporal (millisecond) and spatial (millimeter) resolution by bypassing the signal-dampening effects of skull and scalp. Particular attention is given to high-gamma-power signals (70–200 Hz), which reflect neuronal firing with exceptional signal-to-noise ratios. The chapter addresses methodological considerations including electrode localization, signal processing, and data interpretation challenges unique to recording from epilepsy patients. It balances discussion of iEEG’s remarkable advantages, such as direct access to neuronal activity across cortical layers and network nodes along with its limitations, including restricted accessibility, sparse sampling, and the clinical constraints that dictate electrode placement. The ethical framework governing this invasive research methodology is emphasized throughout.