Lecture 10: Development, Nature and Nurture I

What this lecture covers

This lecture opens the course's unit on brain development by asking the classic nature-versus-nurture question in its modern, empirical form: which parts of the brain's functional organization are built in by genes, and which are shaped by experience? Kanwisher uses face perception as the test case because it has the richest data set, tracing findings from newborn behavior studies through infant and monkey brain imaging.

By the end you'll understand the basic timeline of early brain development, the behavioral evidence that newborns already show sophisticated face-processing abilities, the phenomenon of perceptual narrowing, and why functional MRI studies of face-selective brain regions in monkeys complicate the simple "faces are innate" story. The lecture sits early in the course's development unit and previews the following lecture on space representation.

Key ideas

• Nature versus nurture as an empirical question: Kant argued some structure must be built into the mind prior to experience; the lecture treats this as testable rather than purely philosophical.
• Confound of experience and maturation: as infants age they gain both more experience and more biological maturation, so changes after birth cannot automatically be attributed to learning.
• Preferential looking and habituation: low-tech methods (measuring how long or how far an infant looks) that reveal what infants can discriminate, including Spelke's object-unity experiments.
• Face detection at birth: newborns turn their eyes and heads farther toward schematic face patterns than scrambled or blank patterns, though a simple "more elements on top" cue may explain much of this.
• Perceptual narrowing: young infants can discriminate faces (and phonemes) that adults and older infants can no longer tell apart, because unneeded discriminations get pruned with experience.
• Sugita's controlled-rearing study: monkeys deprived of any face exposure for up to 24 months still showed adult-like face preference and discrimination on first exposure, suggesting strong innate contribution.
• Face patches require face experience: contradicting Sugita, fMRI studies (Livingstone's lab) found that monkeys raised without ever seeing faces show no face-selective patches at all.
• The unresolved puzzle: behavioral and neural evidence about innateness of face perception do not yet converge, which the lecture treats as an open research problem rather than a solved one.

Walkthrough

Framing the nature-nurture question (0:11)

The lecture begins with the philosophical background (Locke and Hume's empiricism versus Kant's a priori structure) before turning it into an empirical question: how does the brain end up with such consistent spatial organization of regions for faces, places, shapes, and colors across individuals? Kanwisher lays out the two ends of a spectrum: minimal innate precursors that bootstrap learning (illustrated with Shimon Ullman's "mover" model for how infants might learn to track hands and gaze from simple motion cues), versus a largely innate, only fine-tuned system.

Basic facts of brain development (5:40)

Most neurons and long-range connections exist at birth, but cortical thickness, neuronal complexity, and myelination all increase rapidly in the first two years. This sets up the key methodological challenge: after birth, maturation and experience are confounded, and later development (like puberty) need not be learned just because it happens later.

Face perception at birth (9:41)

Using paddle-following experiments, newborns within an hour of birth track schematic faces farther than scrambled patterns, though some researchers argue a simple "more elements on top" template accounts for the effect. Using Spelke's habituation-of-looking-time method, Kanwisher explains how researchers infer what infants perceive behind an occluder, then applies the same logic to face recognition studies from an Italian lab: 1- to 3-day-old infants can discriminate individual faces across some viewpoint changes and show an inversion effect for internal facial features, suggesting some face-specific processing very early on.

Perceptual narrowing (31:31)

A live class demo shows that adults can no longer discriminate two monkey faces that 6-month-old infants could tell apart. This perceptual narrowing also happens with speech phonemes and other-race faces, and appears to be reversible only within a developmental window (roughly up to age 12) and only with genuine social interaction, not passive exposure like a video.

Controlled rearing and the Sugita paradox (48:56)

Sugita's study reared monkeys without any face exposure for up to 24 months; on first exposure the monkeys still showed adult-like face preference and discrimination, then went through perceptual narrowing for the face types they weren't shown. This is presented as strong evidence for an innate contribution, though it has never been replicated.

Brain imaging contradicts the behavioral story (56:36)

Rebecca Saks's infant fMRI work shows an adult-like spatial layout of face- and scene-selective regions already by 6 months. But Margaret Livingstone's monkey studies find no face-selective patches until roughly 170 days after birth in normally reared monkeys, and face-deprived monkeys scanned at 260 days show no face patches at all, arguing that "seeing faces is necessary for face domain formation." Kanwisher walks through why this doesn't flatly contradict Sugita (different measures, behavior versus brain), while acknowledging the field's findings are still in tension.

Open questions and discussion (1:03:12)

The lecture closes with unresolved questions: how do face patches know where to develop in the brain, could face-specific neurons exist at birth but be spatially scattered before clustering, and what does perceptual narrowing tell us about metabolic efficiency in the brain. Kanwisher fields student questions on other-race effects, bilingual phoneme retention, and the reliability of fMRI as a measure of underlying neural activity.

Before you watch

• Review the earlier lectures on face-selective regions (the FFA and OFA) and the concept of the inversion effect, since this lecture assumes familiarity with both.
• Read the Sugita controlled-rearing paper referenced throughout, as the lecture builds directly on it.
• Recall the concept of invariance in visual recognition from earlier in the course, since it reappears when discussing face recognition across viewpoints.

Check your understanding

1. Why can't researchers conclude that an ability is learned just because it appears later in development?
2. What did the Sugita study find, and why did it surprise the class?
3. How do the fMRI findings on monkey face patches complicate the Sugita results?
4. What is perceptual narrowing, and what does it suggest about how experience shapes perception?
5. What kind of social experience seemed necessary for infants to retain the ability to discriminate monkey faces or foreign phonemes?