Lecture 1: Introduction to the Human Brain

What this lecture covers

The lecture opens the course with a true story about Kanwisher's friend "Bob," who lost his navigational ability years before doctors found a slow-growing tumor next to his parahippocampal place area, a brain region Kanwisher's own lab had discovered. The story sets up the course's central claim: the brain has structure, some parts of it do remarkably specific jobs, and losing one part can leave the rest of the mind intact.

After the story, Kanwisher explains why the course studies the brain (self-knowledge, the limits of human knowledge, comparison with AI, and what she calls the greatest intellectual quest) and how it will study it, through cognitive science methods paired with brain measurements. She closes with course logistics: grading, readings, and how to read a scientific paper. After watching, you'll know the course's approach, its main themes, and what to expect from the syllabus.

Key ideas

• Functional specificity: some brain regions carry out very specific mental functions, so damage to one region can wipe out a narrow ability while leaving intelligence and personality intact.
• The organization of the brain echoes the architecture of the mind: studying the brain's parts is a way of discovering what the fundamental components of the mind are.
• Parahippocampal place area (PPA): a region Kanwisher's lab found responds selectively to scenes, not faces or objects, and sits near the brain region damaged in Bob's case.
• Neuropsychology as method: testing what a person with brain damage can and cannot do (as Kanwisher did informally with Bob's drawings) is one of the oldest ways to map mental functions to brain regions.
• AI versus human cognition: deep networks like AlexNet now rival humans on standard image classification, but they struggle with harder, more variable images and lack the world models that let people understand a scene, not just label it.
• Multiple methods, multiple levels: behavioral observation, structural brain imaging, functional MRI, and studies of brain damage each reveal a different kind of information about the brain.
• Course focus: the class centers on cognitive neuroscience, meaning mental functions and their brain basis, not molecular or circuit-level neuroscience, and leans heavily on high-level vision and audition.

Walkthrough

The story of Bob (0:10)

Kanwisher tells an extended personal story about a close friend who collapsed one morning, was found to have a large mass in his brain, and underwent major surgery. Years earlier, Bob had shown subtle signs of getting lost in familiar places, which Kanwisher had noticed but not connected to a medical problem. Testing him informally before his surgery (asking him to sketch floor plans and draw objects), she found he could not reproduce spatial layouts but could draw multi-part objects like a bicycle normally, suggesting a specific deficit in spatial memory rather than a general drawing or memory problem. After surgery, his navigational ability never returned, illustrating that recovery from damage to specialized circuits is often limited in adults.

Themes drawn from the story (28:36)

Kanwisher pulls several recurring themes from the story: the brain has structure rather than being uniform mush, some regions are highly specific in function, the organization of the brain reflects the organization of the mind, brains change differently depending on age and type of damage, and many different methods (behavior, anatomical imaging, functional imaging, and studies of preserved and lost abilities) are needed to understand it.

Why study the brain (31:39)

She gives four reasons: understanding your own identity, since the brain (unlike other organs) is who you are; understanding the limits of human knowledge; advancing and comparing with AI; and, most compelling to her, pursuing what she calls the greatest intellectual quest. The AI discussion covers AlexNet's 2012 breakthrough on ImageNet, its human-like error patterns, and later work showing deep nets perform much worse than humans on more variable, realistic images, plus examples of image-captioning systems that label objects correctly but miss the meaning of a scene.

How the course will study the brain (45:47)

Kanwisher explains the course will ask how the brain gives rise to the mind, starting from mental functions like perception and cognition rather than from molecules, neurons, or circuits. For each function, the course will ask whether it has specialized brain machinery and what information is represented, using methods from cognitive science (psychophysics, illusions) and neuroscience (patient studies, fMRI, single-neuron recording, EEG, MEG, and tractography).

Course scope and topics (53:50)

She lists the mental functions the course will cover in depth, including color, shape and motion perception, face, place, and word recognition, number cognition, speech and music perception, language, and theory of mind, and explains why topics like motor control, subcortical function, decision making, and detailed memory research are left out. She contrasts the field's progress from 1990, when only a handful of regions had known functions, to now, when dozens of regions are reasonably well understood.

Logistics and how to read a paper (1:02:55)

The lecture ends with grading breakdown (midterm, final, weekly reading responses, quizzes, and a longer experiment-design assignment) and a walkthrough of how to read a primary research article: find the question in the abstract, find the result, find the interpretation, then dig into the methods and design, while skipping technical detail that isn't central to the course.

Before you watch

• No prior knowledge is assumed; this is the first lecture of the course.
• Familiarity with the concept of brain regions having distinct functions (e.g. language areas, visual cortex) is helpful but not required.
• If you have taken MIT 9.00 or 9.01, expect light overlap with this introduction.

Check your understanding

1. Why did Bob's IQ and personality remain intact despite losing a specific mental ability?
2. What did the comparison between Bob's earlier and later brain scans reveal about the growth rate of his tumor, and why did that matter clinically?
3. According to the lecture, why do deep networks like AlexNet perform far worse on Boris Katz and Andrei Barbu's more variable image set than on standard ImageNet images?
4. What does Kanwisher mean when she says "the organization of the brain echoes the architecture of the mind"?
5. What is the first thing you should look for when reading a scientific paper, according to the lecture's advice?