Introduction to Behavioral Neuroscience · Comparative Neuroscience
How Do We Compare Brains?
On this page 9 sections
In 30 seconds
Once a species is chosen as a model system, the next step is often comparing its brain with others'. Comparative neuroanatomy compares brains at many levels: gross size and shape, regional proportions, neuron numbers, layering and folding, connectivity, and gene expression. The guiding insight: brains are not simply scaled-up versions of one another — lineages evolved different solutions to the same problems, and even within mammals, brain size, neuron density, and organization vary in ways that do not track body size or intelligence simply.
This topic gives you the tools to read comparative claims critically: what allometry How a body part's size scales with body size (brain grows slower) Full entry → is, why "bigger brain equals smarter" is a myth, and what EQ, neuron counts, homology Structures shared because of common ancestry Full entry →, and analogy Structures similar in function but with independent origins Full entry → really measure.
Why this matters
Comparative brain data shape how we think about human uniqueness, intelligence, and disease. Claims like "whales have bigger brains, so why aren't they smarter?" only make sense with the concepts here. Scaling and neuron counts matter for interpreting research on human brain evolution, popular-science headlines, and which animal models suit human brain disorders.
The college version
Core Concepts
Levels of comparison
Brains can be compared at several scales, each answering a different question:
- Gross anatomy: total mass/volume and relative size of major parts (cortex, cerebellum, olfactory bulbs, hippocampus).
- Cytoarchitecture: cell types, layering (the six layers of mammalian neocortex Six-layered cortex characteristic of mammals (commonly taught) Full entry →, commonly taught), neuron density.
- Connectivity: which regions connect to which (tract tracing, diffusion imaging).
- Molecular level: gene expression, receptor distributions, neuromodulator systems.
Allometry: body size is the confound
Brain size scales with body size sub-linearly: brains grow more slowly than bodies (a power law, usually plotted on log–log axes), so raw masses cannot be compared without accounting for body size. The encephalization quotient (EQ) Actual brain size ÷ predicted brain size Full entry → does that: actual brain size divided by the size predicted for the body. Humans have a notably high EQ — among the highest of any mammal, commonly taught — while shrews have small absolute brains but high relative investment. EQ is a rough index, not a measure of intelligence, but it corrects the biggest error in raw-mass comparisons.
Neuron counts: grams are not neurons
Brain mass can mislead because brains differ in neuron density. The isotropic fractionator Dissolves brain tissue and counts stained cell nuclei Full entry → method (dissolving brain tissue and counting stained nuclei) produced striking results (commonly cited estimates; check current literature):
- The human brain contains roughly 86 billion neurons, about 16 billion in the cerebral cortex.
- An African elephant brain is far heavier (~5 kg vs. ~1.4 kg for a human) with more total neurons (~257 billion), but the vast majority sit in the cerebellum, not the cortex.
- The human cerebral cortex is unusually rich in neurons for a primate-sized brain.
The lesson: total mass and neuron count Estimated total neurons, often via the isotropic fractionator Full entry → are poor proxies for "cortical processing power" — where neurons sit, and how densely, matters as much as how many there are.
Homology versus analogy
Distinguish homologous structures — shared through common ancestry (all vertebrate forebrains) — from analogous ones, similar in function but evolved independently (octopus and vertebrate camera eyes are convergent, not homologous). Homology justifies generalizing findings between species; analogy warns that similar function need not imply shared mechanism. Comparative studies must also correct for shared ancestry: species are not independent data points, so naive cross-species correlations can be spurious (phylogenetic correction Statistics accounting for shared ancestry in cross-species comparisons Full entry → methods address this).
Cortical organization: not all brains are built alike
- Mammals: the hallmark is the six-layered neocortex (commonly taught). Many small mammals (rodents, shrews) have lissencephalic Smooth, unfolded cortical surface (rodents) Full entry → (smooth) brains; primates, whales, and elephants have gyrencephalic Folded cortical surface (primates, whales) Full entry → (folded) brains — folding adds surface area within the skull.
- Birds: birds lack a six-layered neocortex but have the hyperpallium and dorsal ventricular ridge. For decades this was used to argue birds were cognitively simple — the "bird brain" insult. Modern work shows corvids and parrots rival primates in tool use, planning, and social reasoning with a different architecture. Complex cognition does not require a mammalian cortex.
Regional specializations reveal behavioral priorities
Brain-region proportions and maps track each species' sensory and motor demands: primates invest heavily in visual cortex, rodents devote large somatosensory territory to the barrel cortex (whiskers), dogs have large olfactory bulbs, and the cerebellum is enlarged in electric fish and humans. Maps (like the cortical homunculus) are distorted: area reflects behavioral importance, not body size — hence humans' huge hand and face representations.
Caveats: methods and interpretation
Brain mass changes with hydration, fixation, and age; neuron counts are estimates depending on method and sampling; and "intelligence" is not one measurable variable. Across-species correlations between brain size and cognition are real but weak and can reverse with different measures. Always ask: compared at which level, corrected for what, measured how?
Common Confusions
| Do Not Confuse | With | Difference |
|---|---|---|
| "Bigger brain = smarter" | EQ, neuron counts, organization | Whales and elephants have larger brains but different neuron distributions and EQs |
| "All brains are scaled-up versions of each other" | Different architectures per lineage | Birds achieve complex cognition without a neocortex; folding and density vary |
| "Homology" | Analogy | Homology = shared ancestry; analogy = convergent function (octopus vs. human eye) |
| "The neocortex exists in all vertebrates" | Six-layered neocortex is mammalian | Birds, fish, and reptiles lack it yet show sophisticated behavior |
| "Neuron counts are exact, fixed numbers" | Method-dependent estimates | Fractionator vs. stereology and sampling change them |
| "Brain mass from a specimen is a clean metric" | Hydration/fixation and body size affect it | Mass needs correction (EQ) and method caveats |
Eli explains
The same idea, in plain words
Explain it like I’m 10
Comparing brains is like comparing cars: you don't just weigh them — you look at the engine, how parts are arranged, and what the car does. A heavy truck and a light sports car are different machines; a big brain and a smart brain are different questions.
Worked example
Compare three mammals: a mouse brain weighs about half a gram, a human brain about 1.4 kg, an African elephant brain about 5 kg (commonly cited reference values). Ranked by mass, the elephant wins — but add the other levels: correcting for body size, the elephant's EQ is well below the human's; counting neurons, the elephant has more total neurons (~257 billion), but most are in the cerebellum, while the human cerebral cortex has more (~16 billion vs. ~5.6 billion in the elephant cortex, commonly cited estimates). If cognition depended on cortical neurons, the human cortex's density — not the elephant's mass — is the striking feature; and a bird with no neocortex can outperform many mammals on planning tasks. Mass, EQ, neuron count, and architecture each tell a different part of the story.
Key takeaways
- Brains are not scaled-up versions of one another; lineages evolved different architectures (mammals vs. birds).
- Brain mass scales sub-linearly with body mass (allometry); EQ corrects for body size, and humans score notably high.
- Mass and neuron count can mislead: the elephant brain is heavier (~5 kg) with more total neurons (~257 billion), but most are cerebellar; the human cortex has more (~16 billion, commonly cited).
- The six-layer neocortex is mammalian; birds achieve complex cognition with the hyperpallium/DVR — the "bird brain" myth is false.
- Homologous = shared ancestry; analogous = convergent function; correct statistically for shared ancestry.
- Folding (gyrencephaly) adds cortical surface area; smooth (lissencephalic) brains can still be efficient.
- Map sizes (e.g., human hand/face in the homunculus) reflect behavioral importance, not body size.
- Neuron counts and brain masses are estimates — treat all values as approximate.
Check yourself
6 review questions from the chapter. Try each one, then open the answer.
Why can't you compare raw brain masses across species?
Show answer
Because brain size scales with body size (allometry); comparisons need correction (e.g., EQ).
What does the encephalization quotient (EQ) correct for, and which species is famous for a high EQ?
Show answer
EQ corrects for body size — actual brain size divided by the predicted size for the body; humans are commonly cited among the highest.
Using the commonly cited numbers, why does the elephant's larger brain not imply more cortical processing power than a human's?
Show answer
The elephant brain is heavier (~5 kg vs. ~1.4 kg) with more total neurons (~257 billion), but most are cerebellar; the human cortex has more neurons (~16 billion vs. ~5.6 billion, commonly cited) — cortical count differs, not total mass.
What is the difference between homologous and analogous structures?
Show answer
Homologous structures share common ancestry (vertebrate forebrains); analogous structures are similar in function but evolved independently (octopus and vertebrate eyes).
Why is calling birds "bird-brained" scientifically wrong?
Show answer
Because birds lack a mammalian neocortex yet show cognition rivaling primates (corvids, parrots) using the hyperpallium/DVR — complex cognition needs no neocortex.
What is the isotropic fractionator used for, and why are its outputs called estimates?
Show answer
It dissolves brain tissue and counts stained nuclei to estimate total neurons; outputs depend on method, sampling, and tissue handling.
Study tools & related lessonsKey vocabulary · Related
Key vocabulary
- allometry
- How a body part's size scales with body size (brain grows slower)
- encephalization quotient (EQ)
- Actual brain size ÷ predicted brain size
- neuron count
- Estimated total neurons, often via the isotropic fractionator
- isotropic fractionator
- Dissolves brain tissue and counts stained cell nuclei
- lissencephalic
- Smooth, unfolded cortical surface (rodents)
- gyrencephalic
- Folded cortical surface (primates, whales)
- homology
- Structures shared because of common ancestry
- analogy
- Structures similar in function but with independent origins
- neocortex
- Six-layered cortex characteristic of mammals (commonly taught)
- phylogenetic correction
- Statistics accounting for shared ancestry in cross-species comparisons
Sources & references
This lesson was adapted from the open educational references above; their licenses and attributions are preserved. See Copyright & Licensing.
Educational content only. It is not medical, legal or professional advice. Found an error? Tell us.