Cognitive Psychology Series Part 7: Cognitive Neuroscience

1. History of Neuroscience

The quest to understand the brain's role in cognition stretches back millennia, but the path from ancient speculation to modern neuroscience is a story of dramatic paradigm shifts, serendipitous discoveries, and persistent misconceptions.

1.1 From Phrenology to Localization

In the early 19th century, Franz Joseph Gall proposed phrenology — the idea that different mental faculties were localized in specific brain regions and that the size of each region could be assessed by feeling the bumps on a person's skull. While phrenology was fundamentally wrong in its methods (skull bumps tell us nothing about brain function), Gall's core insight — that different brain areas serve different functions — was actually correct and revolutionary for its time.

The first compelling evidence for cerebral localization came from Paul Broca in 1861. His patient "Tan" (Louis Victor Leborgne) could understand speech perfectly but could only produce one syllable: "tan." After Tan's death, Broca's autopsy revealed damage to the left inferior frontal gyrus — now known as Broca's area. Shortly after, Carl Wernicke (1874) identified a region in the left temporal lobe (Wernicke's area) whose damage produced fluent but meaningless speech — patients could produce grammatically correct sentences that made no sense.

1.2 The Neuron Doctrine

Before the late 19th century, scientists debated whether the brain was a continuous network (reticular theory, championed by Camillo Golgi) or composed of discrete cells. Santiago Ramon y Cajal, using Golgi's own staining technique, demonstrated that the nervous system was made of individual cells — neurons — separated by tiny gaps (later called synapses by Charles Sherrington). Cajal and Golgi shared the 1906 Nobel Prize, though they disagreed fundamentally about what they had discovered.

2. Brain Regions & Cognitive Functions

The human cerebral cortex is divided into four lobes, each with specialized but overlapping functions. Beneath the cortex lie subcortical structures — the hippocampus, amygdala, basal ganglia, and cerebellum — that are critical for memory, emotion, habit formation, and motor coordination.

2.1 The Prefrontal Cortex — CEO of the Brain

The prefrontal cortex (PFC), located behind your forehead, is the most recently evolved brain region and the last to fully mature (not until your mid-20s). It is the seat of executive functions — the higher-order cognitive abilities that make us uniquely human:

• Planning and decision-making: Evaluating options, weighing consequences, and selecting actions
• Working memory: Holding and manipulating information in mind (dorsolateral PFC)
• Inhibitory control: Suppressing inappropriate responses and impulses (ventrolateral PFC)
• Cognitive flexibility: Switching between tasks and mental sets
• Social cognition: Understanding others' mental states, moral reasoning (medial PFC)
• Personality and emotional regulation: Integrating emotion with rational thought (orbitofrontal cortex)

2.2 The Hippocampus & Amygdala

Deep within the temporal lobes lie two structures critical for memory and emotion:

The hippocampus (Greek for "seahorse," reflecting its shape) is essential for forming new declarative memories. As we learned in Part 1, patient H.M.'s bilateral hippocampal removal left him unable to form new episodic or semantic memories. The hippocampus acts as a memory indexer — it doesn't store memories permanently but binds together the various cortical regions that encode different aspects of an experience (sights, sounds, emotions, context) into a coherent memory trace. Over time, through systems consolidation, these connections strengthen enough that the cortex can retrieve the memory independently.

The amygdala (almond-shaped, located adjacent to the hippocampus) is the brain's emotional sentinel. It is especially critical for:

• Fear conditioning: Learning to associate stimuli with danger (crucial for survival)
• Emotional memory enhancement: Emotionally arousing events are remembered better because the amygdala modulates hippocampal consolidation
• Threat detection: The amygdala can process potential threats before conscious awareness via a fast subcortical pathway (LeDoux's "low road")
• Social evaluation: Judging trustworthiness of faces, recognizing emotional expressions

2.3 The Four Cerebral Lobes

2.4 The Cerebellum & Basal Ganglia

The cerebellum ("little brain") sits at the back of the skull and contains more neurons than the rest of the brain combined (approximately 69 billion of the brain's 86 billion neurons). Traditionally viewed as purely a motor structure, modern research has revealed its critical role in:

• Motor coordination: Timing, precision, and smoothness of movements
• Motor learning: Adapting movements through practice (procedural memory)
• Cognitive timing: Internal clock functions, rhythm perception
• Language processing: Articulatory rehearsal, verbal working memory
• Emotional regulation: The "cerebellar cognitive affective syndrome" after cerebellar damage

The basal ganglia are a group of subcortical nuclei (caudate, putamen, globus pallidus, subthalamic nucleus, substantia nigra) that form critical loops with the cortex. They are essential for:

• Habit formation: Transitioning from goal-directed to automatic behavior
• Reward-based learning: The dopaminergic reward prediction error signal
• Action selection: Choosing which action to execute and suppressing competing alternatives
• Procedural memory: Learning sequences and routines

3. Neural Networks & Connectivity

Modern neuroscience has shifted from a "brain region" approach to a network approach. Cognition emerges not from individual brain areas but from the dynamic interactions between distributed networks. Three large-scale networks are particularly important for understanding cognition:

3.1 The Default Mode Network (DMN)

Discovered by Marcus Raichle in 2001, the Default Mode Network is active when you are not focused on the external world — during mind-wandering, daydreaming, thinking about yourself, remembering the past, or imagining the future.

Key regions: Medial prefrontal cortex, posterior cingulate cortex, precuneus, angular gyrus, lateral temporal cortex, hippocampal formation.

Functions:

• Self-referential thought ("What kind of person am I?")
• Autobiographical memory retrieval
• Future planning and mental simulation
• Theory of mind (understanding others' perspectives)
• Creative incubation and insight

3.2 The Salience Network

The Salience Network acts as a switch operator, detecting important (salient) stimuli and toggling the brain between the internally-focused DMN and the externally-focused Central Executive Network.

Key regions: Anterior insula, dorsal anterior cingulate cortex (dACC).

Functions:

• Detecting novel, unexpected, or emotionally significant stimuli
• Switching between internal and external attention
• Integrating sensory, emotional, and cognitive information
• Error monitoring and conflict detection
• Autonomic regulation (heart rate, arousal)

3.3 The Central Executive Network (CEN)

The Central Executive Network engages during cognitively demanding tasks that require focused attention, working memory, and deliberate decision-making.

Key regions: Dorsolateral prefrontal cortex (dlPFC), posterior parietal cortex (PPC).

Functions:

• Working memory maintenance and manipulation
• Sustained attention to external tasks
• Logical reasoning and problem-solving
• Cognitive control and inhibition

4. Brain Imaging Techniques

The ability to observe the living brain in action has revolutionized cognitive neuroscience. Each imaging technique offers a different window into brain function, with distinct trade-offs between spatial resolution (how precisely we can localize activity) and temporal resolution (how precisely we can track when activity occurs).

4.1 Functional Magnetic Resonance Imaging (fMRI)

fMRI is the workhorse of modern cognitive neuroscience. It works by detecting changes in blood oxygenation — the BOLD (Blood-Oxygen-Level-Dependent) signal. When neurons become active, they consume oxygen; the brain overcompensates by sending extra oxygenated blood to the active area. fMRI detects this hemodynamic response.

Strengths: Excellent spatial resolution (~1-3mm), non-invasive, no radiation, can image deep brain structures.

Limitations: Poor temporal resolution (~1-2 seconds — the hemodynamic response lags behind neural activity by several seconds), expensive equipment, sensitive to head movement, indirect measure of neural activity.

4.2 Electroencephalography (EEG)

EEG records electrical activity from the scalp using arrays of electrodes. It directly measures the summed postsynaptic potentials of thousands of neurons firing synchronously.

Strengths: Excellent temporal resolution (~1 millisecond), inexpensive, portable, direct measure of neural electrical activity, excellent for studying Event-Related Potentials (ERPs).

Limitations: Poor spatial resolution (~1-3 cm), cannot image deep brain structures, signals distorted by skull and scalp tissue.

4.3 PET, MEG & TMS

# Simulating EEG signal processing and Event-Related Potential extraction
import numpy as np

class EEGProcessor:
    """
    Simulates basic EEG signal processing for cognitive neuroscience.
    Demonstrates filtering, epoching, and ERP extraction.
    """

    def __init__(self, sampling_rate=256, duration_sec=10):
        self.fs = sampling_rate
        self.duration = duration_sec
        self.n_samples = self.fs * self.duration
        self.time = np.linspace(0, self.duration, self.n_samples)

    def generate_raw_eeg(self):
        """Generate simulated raw EEG with multiple frequency bands."""
        # Alpha waves (8-13 Hz) - relaxed wakefulness
        alpha = 10 * np.sin(2 * np.pi * 10 * self.time)
        # Beta waves (13-30 Hz) - active thinking
        beta = 5 * np.sin(2 * np.pi * 20 * self.time)
        # Theta waves (4-8 Hz) - drowsiness, memory encoding
        theta = 7 * np.sin(2 * np.pi * 6 * self.time)
        # Delta waves (0.5-4 Hz) - deep sleep
        delta = 15 * np.sin(2 * np.pi * 2 * self.time)
        # Random noise
        noise = np.random.normal(0, 3, self.n_samples)

        raw_signal = alpha + beta + theta + delta + noise
        return raw_signal

    def bandpass_filter(self, signal, low_freq, high_freq):
        """Simple bandpass filter using FFT."""
        fft_signal = np.fft.fft(signal)
        frequencies = np.fft.fftfreq(len(signal), 1.0 / self.fs)

        # Zero out frequencies outside the passband
        mask = (np.abs(frequencies) >= low_freq) & (np.abs(frequencies) <= high_freq)
        filtered_fft = fft_signal * mask
        filtered_signal = np.real(np.fft.ifft(filtered_fft))
        return filtered_signal

    def extract_frequency_bands(self, signal):
        """Extract standard EEG frequency bands."""
        bands = {
            'Delta (0.5-4 Hz)': self.bandpass_filter(signal, 0.5, 4),
            'Theta (4-8 Hz)': self.bandpass_filter(signal, 4, 8),
            'Alpha (8-13 Hz)': self.bandpass_filter(signal, 8, 13),
            'Beta (13-30 Hz)': self.bandpass_filter(signal, 13, 30),
            'Gamma (30-100 Hz)': self.bandpass_filter(signal, 30, 100)
        }
        return bands

    def compute_band_power(self, signal):
        """Compute power spectral density for each frequency band."""
        bands = self.extract_frequency_bands(signal)
        power = {}
        for band_name, band_signal in bands.items():
            power[band_name] = np.mean(band_signal ** 2)
        return power

    def simulate_erp(self, n_trials=100, stimulus_onset=0.2):
        """
        Simulate Event-Related Potential (ERP) extraction.
        ERPs emerge from averaging many trials to cancel out noise.
        """
        epoch_duration = 0.8  # 800ms epochs
        n_epoch_samples = int(epoch_duration * self.fs)
        epochs = []

        for trial in range(n_trials):
            # ERP components (signal)
            t_epoch = np.linspace(0, epoch_duration, n_epoch_samples)
            # N100 component (~100ms post-stimulus, negative)
            n100 = -8 * np.exp(-((t_epoch - 0.1) ** 2) / (2 * 0.015 ** 2))
            # P200 component (~200ms, positive)
            p200 = 6 * np.exp(-((t_epoch - 0.2) ** 2) / (2 * 0.02 ** 2))
            # P300 component (~300ms, positive — attention/decision)
            p300 = 10 * np.exp(-((t_epoch - 0.35) ** 2) / (2 * 0.04 ** 2))
            # N400 component (~400ms, negative — semantic processing)
            n400 = -5 * np.exp(-((t_epoch - 0.45) ** 2) / (2 * 0.03 ** 2))

            signal = n100 + p200 + p300 + n400
            noise = np.random.normal(0, 15, n_epoch_samples)
            epochs.append(signal + noise)

        # Average across trials (noise cancels out, signal remains)
        erp = np.mean(epochs, axis=0)

        print("=== ERP Extraction Results ===")
        print(f"Number of trials averaged: {n_trials}")
        print(f"Signal-to-noise improvement: {np.sqrt(n_trials):.1f}x")
        print(f"\nERP Components Detected:")
        print(f"  N100 (sensory): ~100ms, amplitude = {np.min(erp[:int(0.15*self.fs)]):.1f} uV")
        print(f"  P200 (attention): ~200ms, amplitude = {np.max(erp[int(0.15*self.fs):int(0.25*self.fs)]):.1f} uV")
        print(f"  P300 (decision): ~350ms, amplitude = {np.max(erp[int(0.25*self.fs):int(0.4*self.fs)]):.1f} uV")
        print(f"  N400 (semantic): ~450ms, amplitude = {np.min(erp[int(0.4*self.fs):]):.1f} uV")

        return erp

# Demo
processor = EEGProcessor(sampling_rate=256, duration_sec=10)
raw = processor.generate_raw_eeg()
power = processor.compute_band_power(raw)

print("=== EEG Band Power Analysis ===")
for band, p in power.items():
    bar = "█" * int(p / 5)
    print(f"  {band:25s}: {p:8.2f} uV^2  {bar}")

print()
erp = processor.simulate_erp(n_trials=100)

5. Neuroplasticity

Neuroplasticity — the brain's ability to reorganize its structure and function in response to experience — is arguably the most important discovery in modern neuroscience. It overturned the long-held belief that the adult brain was fixed and unchangeable, revealing instead that the brain is a dynamic organ that continuously rewires itself throughout life.

5.1 Hebb's Rule & Long-Term Potentiation (LTP/LTD)

In 1949, Donald Hebb proposed a deceptively simple principle that became the foundation of learning theory in neuroscience: "Neurons that fire together, wire together." More formally, when neuron A repeatedly participates in firing neuron B, the connection between them is strengthened.

The biological mechanism underlying Hebb's rule was discovered in 1973 by Bliss and Lomo, who demonstrated Long-Term Potentiation (LTP) in the hippocampus — a persistent increase in synaptic strength following high-frequency stimulation. LTP has become the leading cellular model of learning and memory.

5.2 Critical Periods & Adult Neurogenesis

Critical periods are windows of heightened plasticity during early development when the brain is particularly sensitive to environmental input. The most famous example is the visual system: if one eye is deprived of input during the critical period (roughly the first few years of life in humans), the visual cortex permanently reallocates neurons to the functional eye. After the critical period closes, this reorganization cannot be reversed.

Key examples of critical periods:

• Vision: Binocular vision requires coordinated input from both eyes during the first years of life (Hubel & Wiesel, Nobel Prize 1981)
• Language: Native-like phonetic discrimination narrows by age 10-12 months; grammar acquisition is optimal before puberty (Lenneberg's Critical Period Hypothesis)
• Attachment: Secure attachment bonds form primarily in the first 2-3 years

For decades, scientists believed that no new neurons were generated in the adult brain. This dogma was shattered when adult neurogenesis was definitively demonstrated in the hippocampus (Eriksson et al., 1998). New neurons are continuously produced in the dentate gyrus of the hippocampus and the subventricular zone throughout adulthood. Factors that enhance adult neurogenesis include aerobic exercise, enriched environments, learning, and adequate sleep. Chronic stress and depression suppress it.

5.3 London Taxi Drivers — Neuroplasticity in Action

6. Brain Injuries & Cognition

Some of the most profound insights into brain function have come from studying people whose brains were damaged by accident, disease, or surgery. These "experiments of nature" reveal what specific brain regions contribute to cognition by showing what happens when they are removed from the circuit.

6.1 Phineas Gage — The Most Famous Brain Injury in History

6.2 Split-Brain Patients

In the 1960s, Roger Sperry and Michael Gazzaniga studied patients who had undergone corpus callosotomy — surgical severing of the corpus callosum (the 200+ million nerve fibers connecting the two hemispheres) — to treat severe epilepsy. These "split-brain" patients provided extraordinary insights into hemispheric specialization.

Key findings from split-brain research:

6.3 Traumatic Brain Injury (TBI)

Traumatic Brain Injury affects approximately 69 million people worldwide each year. The cognitive consequences depend on the location, severity, and type of injury:

Chronic Traumatic Encephalopathy (CTE) — a progressive neurodegenerative disease caused by repeated head impacts — has gained significant attention due to its prevalence among athletes in contact sports. CTE involves the accumulation of abnormal tau protein and progressive deterioration of memory, executive function, mood regulation, and eventually motor function.

7. Neurotransmitters & Cognition

Neurons communicate across synapses using chemical messengers called neurotransmitters. Different neurotransmitter systems modulate different aspects of cognition, and their dysfunction underlies many cognitive and psychiatric disorders.

Exercises & Self-Assessment

Conclusion & Next Steps

In this chapter, we have journeyed from the earliest attempts to map the mind onto the brain to the cutting-edge network neuroscience of today. Here are the key takeaways:

• The brain is organized into specialized regions — the prefrontal cortex for executive function, the hippocampus for memory, the amygdala for emotion — but cognition emerges from the interaction of distributed networks, not isolated areas
• Three large-scale networks — the Default Mode, Salience, and Central Executive Networks — dynamically interact to produce the full range of cognitive experience, from daydreaming to focused problem-solving
• Brain imaging techniques each offer unique windows into brain function, with fundamental trade-offs between spatial and temporal resolution. Only TMS can establish causal brain-behavior relationships
• Neuroplasticity — from Hebb's rule to adult neurogenesis — means the brain continuously rewires itself in response to experience, challenging fixed notions of intelligence and talent
• Brain injuries like Phineas Gage's and split-brain patients have provided irreplaceable insights into the neural basis of personality, consciousness, and hemispheric specialization
• Neurotransmitters modulate cognition in precise, dose-dependent ways — following an inverted U-shaped function where both deficiency and excess impair performance