The Cattell-Horn-Carroll Theory (CHC) of Intelligence

Most of the IQ tests in serious use today share a hidden blueprint. Whether it is a Wechsler scale, the Woodcock-Johnson battery, or a school cognitive assessment, the way the test is divided into parts usually traces back to one framework: the Cattell-Horn-Carroll theory of intelligence, almost always shortened to CHC theory. It is not a single dramatic idea but a carefully organised map of the many things we mean when we call someone "intelligent". This article walks through what CHC theory claims, where it came from, and why it quietly governs the score report you receive at the end of a modern test.

The problem CHC was built to solve

For most of the twentieth century, researchers argued about the shape of intelligence rather than agreeing on it. One camp, following Charles Spearman, emphasised a single general factor — often called g — that seemed to run through every mental task. Another camp preferred a handful of distinct abilities, and others proposed long lists of separate talents. Each model captured something real, yet they disagreed, used different names for similar ideas, and were hard to compare directly.

CHC theory is best understood as an attempt to end that fragmentation. Rather than crowning one model the winner, it integrates the competing pictures into a single hierarchical map. It keeps the idea of a broad general factor, keeps the idea of a small number of major ability domains, and keeps the idea of many fine-grained skills — then arranges all three levels so they fit together. The result is less a bold new claim than a shared vocabulary that test developers around the world can agree to use.

The three-stratum structure

The backbone of CHC is a three-level, or three-stratum, hierarchy. At the top (Stratum III) sits general ability, the familiar g-factor: a single, very broad capacity that influences performance across almost every cognitive task. If you want to understand why one number can summarise so much, our overview of what the g-factor is explores that idea in detail.

In the middle (Stratum II) sits a set of broad abilities — usually counted as somewhere between eight and sixteen, depending on the version. These are large domains such as reasoning, memory, and visual processing. They correlate with one another, which is part of why g appears at all, yet they are distinct enough to be measured separately.

At the base (Stratum I) sit dozens of narrow abilities: highly specific skills such as the speed of naming familiar objects, or remembering a string of digits. Each narrow ability belongs under one of the broad abilities above it. This nesting — many narrow skills grouped into a few broad domains, all feeding a single general factor — is the core structural claim of the theory.

The broad abilities in plain language

The broad abilities are where CHC becomes genuinely useful, because each one names a familiar kind of thinking. The most commonly discussed include:

• Fluid reasoning (Gf) — solving novel problems you cannot answer from memory, the engine behind puzzles and abstract reasoning tasks.
• Comprehension-knowledge, or crystallized intelligence (Gc) — the breadth and depth of knowledge you have acquired, including vocabulary and general information.
• Short-term and working memory (Gsm) — holding and manipulating information for a few seconds, closely tied to working memory and intelligence.
• Visual processing (Gv) — perceiving, analysing, and mentally transforming shapes and images, the domain behind spatial ability and mental rotation.
• Auditory processing (Ga) — discriminating and working with patterns of sound, important for music and for early reading.
• Long-term storage and retrieval (Glr) — how efficiently you file information away and pull it back later, which is separate from how much you know.
• Processing speed (Gs) — how quickly you carry out simple, well-practised mental tasks under time pressure.
• Reading and writing (Grw) — acquired literacy skills, from decoding words to written expression.
• Quantitative knowledge (Gq) — acquired mathematical knowledge, such as number facts and procedures.

Two of these anchor a classic distinction that predates the rest: fluid reasoning and crystallized knowledge, which we cover in depth in fluid and crystallized intelligence.

A brief history of the Cattell-Horn-Carroll theory

The framework carries three surnames because it was built in three stages. From the 1940s onward, Raymond Cattell proposed that Spearman's g could be split in two: fluid intelligence (Gf), the ability to reason in the moment, and crystallized intelligence (Gc), the store of learned knowledge. His student John Horn then expanded this Gf-Gc model over the following decades, adding broad factors such as visual processing, short-term memory, and processing speed, until the theory described many abilities rather than two.

Working separately, John Carroll carried out a monumental re-analysis of hundreds of datasets gathered across the century. Published in 1993, his survey proposed the three-stratum model: narrow abilities, broad abilities, and a general factor on top. Because Horn's expanded model and Carroll's three-stratum model were so similar, researchers — notably in work associated with Kevin McGrew — merged them around the turn of the century into the combined Cattell-Horn-Carroll theory. The hyphenated name is essentially a handshake between two research traditions that had reached compatible conclusions.

Why CHC matters for modern IQ tests

CHC is not just an academic diagram; it shapes the tests people actually sit. Major batteries — the Wechsler scales (WAIS and WISC), the Woodcock-Johnson, and the Kaufman (KABC) tests — are now organised, wholly or partly, around CHC broad abilities. That design choice has a visible consequence: instead of reporting a single number, a modern test gives you several index scores, each estimating a different broad ability, often with a composite on top.

This is why two people with the same overall IQ can show quite different profiles — one strong in verbal comprehension (Gc) but slower in processing speed (Gs), the other the reverse. The broad-ability structure lets a report describe strengths and weaknesses rather than flattening everything into one figure. If you would like to see the idea in action, you can try a free symbol-based test and read the outcome as a profile rather than a verdict. An online screening like this is meant for interest and self-reflection, not clinical or diagnostic use.

CHC, the g-factor, and fluid versus crystallized ability

CHC sits on top of two long-running debates, and it handles them with deliberate diplomacy. The first is the argument over the g-factor: does a single general intelligence really exist, or is it a statistical by-product of averaging correlated tests? CHC keeps g at the apex, yet many who use the framework treat the broad abilities as the more practically useful level, and some researchers question how much weight the top stratum should carry. The theory is compatible with taking g seriously without insisting the matter is settled.

The second is the fluid-versus-crystallized distinction that started the whole enterprise. Within CHC these are simply two broad abilities among many, but they remain among the most robust and well-replicated, which is why they still appear, in one form or another, in almost every modern test. Reasonable researchers continue to disagree about the exact number of broad abilities, how they shift across the lifespan, and how culturally fair each measure is. CHC's strength is not that it has ended these debates, but that it gives everyone a common map on which to hold them.