Intelligence is a general mental ability for reasoning, problem solving, and learning. That definition, while broadly accepted among psychologists, only scratches the surface. Intelligence also includes the capacity to think abstractly, comprehend complex ideas, adapt to new environments, learn from experience, and plan ahead. What makes intelligence so fascinating, and so debated, is that it doesn’t describe a single skill. It describes a cluster of cognitive abilities that work together, and researchers have spent over a century trying to figure out exactly how they relate to each other.
The General Intelligence Factor
In 1904, the psychologist Charles Spearman noticed something striking: people who scored well on one type of mental test tended to score well on others, even when the tests measured very different skills. A person strong in vocabulary was more likely than chance to also be strong in spatial reasoning or arithmetic. Spearman called the underlying factor behind this pattern “g,” short for general intelligence.
The idea of g remains one of the most robust findings in psychology. When researchers give large groups of people batteries of cognitive tests, scores on different subtests consistently correlate with one another. That doesn’t mean everyone is equally good at everything. It means there’s a common thread running through diverse mental abilities. Think of it like general athleticism: a naturally coordinated person might not be the best at every sport, but they’ll pick up most physical tasks faster than average.
Fluid and Crystallized Intelligence
One of the most useful distinctions in intelligence research separates what you’re born with the capacity to do from what you’ve learned along the way. Psychologist Raymond Cattell proposed this split in 1943, and the framework has held up remarkably well.
Fluid intelligence is your ability to solve novel problems, spot patterns, and reason through unfamiliar situations without relying on prior knowledge. It’s what you use when you encounter a type of puzzle you’ve never seen before. Fluid intelligence rises through childhood and adolescence, peaks in young adulthood, and then gradually declines.
Crystallized intelligence is the knowledge and skill you accumulate through education, experience, and cultural exposure. Vocabulary, historical facts, professional expertise, and knowing how to navigate social situations all fall under this umbrella. Unlike fluid intelligence, crystallized intelligence can keep growing well into adulthood and only begins to decline slowly toward the end of life. This is why an experienced doctor or mechanic can outperform a younger, sharper-minded novice: their stored knowledge compensates for any dip in raw processing speed.
Multiple Intelligences and Broader Models
Not everyone agrees that a single factor captures what it means to be intelligent. Howard Gardner, a developmental psychologist at Harvard, proposed that intelligence isn’t one thing but at least nine distinct types, each relatively independent of the others:
- Verbal-linguistic: sensitivity to words, language, and their rhythms
- Logical-mathematical: abstract and numerical reasoning
- Spatial-visual: the ability to think in images and mentally manipulate objects
- Bodily-kinesthetic: control of physical movement and skillful handling of objects
- Musical: sensitivity to rhythm, pitch, and tone
- Interpersonal: reading and responding to other people’s emotions and motivations
- Intrapersonal: self-awareness and understanding of your own feelings and thought processes
- Naturalist: recognizing and categorizing patterns in the natural world
- Existential: the capacity to grapple with deep questions about meaning, life, and death
Gardner’s theory resonated widely in education because it validated abilities that traditional IQ tests ignore. A gifted dancer or a socially perceptive leader is “intelligent” in ways that don’t show up on a standardized exam. Critics, however, point out that many of Gardner’s intelligences correlate with each other, which brings us back toward something resembling g. The debate is less about who’s right and more about how broadly you want to define the word.
Robert Sternberg offered a middle ground with his triarchic theory, which divides intelligence into three broad types: analytical intelligence (critical thinking and problem solving, what most people call “book smarts”), creative intelligence (generating novel ideas and approaches), and practical intelligence (solving real-world, everyday problems, often called “street smarts” or common sense). Most people intuitively recognize this distinction. You’ve probably met someone who aces exams but struggles with basic life logistics, or someone with no formal education who navigates complex social and business situations with ease.
How Intelligence Is Measured
The most widely used intelligence test for adults is the Wechsler Adult Intelligence Scale, now in its fourth edition. Rather than producing a single number, it measures four broad cognitive domains: verbal comprehension (vocabulary, general knowledge, abstract verbal reasoning), perceptual reasoning (spatial puzzles, pattern recognition), working memory (holding and manipulating information in your head), and processing speed (how quickly you can scan and compare simple visual information). Ten core subtests feed into these four domain scores, which combine into an overall IQ.
IQ scores are designed so that 100 is the average, with about two-thirds of the population falling between 85 and 115. The tests are periodically re-normed because raw scores have historically risen over time, a phenomenon called the Flynn effect. For most of the 20th century, average IQ scores climbed roughly 2 to 4 points per decade, likely driven by improvements in nutrition, education, and environmental complexity. Recent evidence, however, suggests this trend may be reversing. A 2024 study comparing population-representative samples from 2005 and 2024 found declines in the overall g factor, even though scores on individual subtests still rose in some areas. The pattern is inconsistent across countries and cognitive domains, making it one of the more puzzling trends in intelligence research right now.
What Happens in the Brain
Intelligence doesn’t live in one spot in the brain. It depends on a distributed network of regions working together, but some areas matter more than others. A large study of over 11,000 people found that the dorsolateral prefrontal cortex, a region behind your forehead involved in planning, decision-making, and holding information in mind, showed the strongest link to intelligence. People with greater surface area in this region tended to score higher on cognitive tests, and statistical analyses suggested the relationship is causal, not just correlational.
Other regions also contribute. Parts of the brain involved in language processing and areas behind the eyes involved in social and emotional decision-making showed independent causal links to intelligence. Interestingly, the relationship goes both ways: higher intelligence also appears to promote greater cortical thickness in language-related areas, suggesting that using your brain intensively may physically shape it over time.
Genetics, Environment, and the Nature-Nurture Split
Twin and family studies consistently show that about half of the variation in intelligence between people can be attributed to genetic differences. That figure isn’t fixed across the lifespan, though. Heritability is lower in infancy and childhood, when environment plays a larger role, and increases through adolescence and into adulthood. By the time you’re an adult, your genes account for a larger share of the differences between you and your peers than they did when you were five.
This doesn’t mean intelligence is “50% genetic” in any individual person. Heritability is a population-level statistic. It tells you how much of the variation within a group is explained by genetic differences, under the environmental conditions that group experienced. Change the environment dramatically (say, by eliminating childhood malnutrition across a population) and the heritability estimate would shift, because you’ve removed an environmental source of variation. Genes and environment don’t operate independently. They interact constantly, with genetic tendencies influencing what environments a person seeks out, and environments shaping how genetic potential gets expressed.
Human Intelligence vs. Artificial Intelligence
The rise of AI has renewed interest in what makes human intelligence distinctive. Despite their impressive performance on specific tasks, AI systems operate on a fundamentally different basis than biological brains. They are, at their core, digital machines optimized for particular functions: processing language, recognizing images, playing chess. They can outperform humans dramatically in domains that require exact calculation and rapid data processing.
Human intelligence, by contrast, is general-purpose and deeply tied to consciousness, embodied experience, and social understanding. Humans excel at transferring knowledge between unrelated domains, understanding context and nuance, and navigating ambiguous social situations. Our biological neural networks also come with built-in distortions, though. Cognitive biases like anchoring (over-relying on the first piece of information you encounter), confirmation bias (favoring information that supports what you already believe), and hindsight bias (perceiving past events as more predictable than they were) are side effects of the shortcuts our brains evolved to make fast decisions in a world that looked very different from the one we live in now. AI systems don’t suffer from these particular biases, though they have their own limitations and blind spots shaped by training data and design choices.
For the foreseeable future, AI and human intelligence remain fundamentally different in kind, not just degree. No current AI system possesses the flexible, self-aware, context-sensitive cognition that defines human thought.