Once psychologists began taking general intelligence seriously, another problem became harder to ignore. Cognitive tests may all be positively related, but they are not related in exactly the same way. Verbal tests resemble other verbal tests especially strongly. Spatial tasks cluster with spatial tasks. Numerical, memory and reasoning measures can form recognizable groupings of their own. Louis Thurstone thought those patterns deserved more than the status of leftovers beneath g. His theory of primary mental abilities shifted attention toward several broad dimensions of cognitive performance. The classic formulation included verbal comprehension, word fluency, number facility, spatial ability, associative memory, perceptual speed and reasoning. Thurstone did not ultimately make general covariance disappear. Instead, his work helped expose the level between one overall factor and individual tasks—the broad abilities that later became central to hierarchical theories of intelligence.
Why one general factor seemed insufficient
Spearman’s framework emphasized a general factor common to diverse cognitive performances plus influences specific to individual tests. Thurstone believed this representation was too restrictive for the multidimensionality of mental abilities. Instead of assuming one common factor in advance, he developed and applied multiple-factor methods to large batteries of cognitive tests and asked what recurring dimensions emerged from their correlation patterns. The dispute was therefore partly substantive and partly methodological. Spearman’s approach asked whether correlations could be represented by one common factor plus test-specific influences. Thurstone wanted a method capable of recovering several common dimensions when the data demanded them. His idea of simple structure and the rotation of factor axes were meant to produce patterns in which groups of tests had clearer psychological interpretations. This history matters because factor analysis does not hand researchers a uniquely labeled set of mental faculties. The orientation of factors, the constraints imposed on them, and whether they are allowed to correlate affect the representation. Thurstone’s work helped make those choices visible. It also helped transform the argument from “one factor or many?” into a harder question: at what level of generality should cognitive structure be described? The best-known result was a set of primary mental abilities.
What were the primary mental abilities?
The classic list is commonly summarized as:
- verbal comprehension;
- word fluency;
- number facility;
- spatial ability;
- associative memory;
- perceptual speed;
- reasoning.
The exact labels and factor solutions varied somewhat across studies. The important theoretical move was not the magical discovery of exactly seven permanent mental modules. It was the claim that:
cognitive performance contains broad, recurring dimensions that deserve representation in their own right.
What broad abilities look like in data
Imagine six fictional tests:
- two language/comprehension measures;
- two spatial measures;
- two reasoning measures.
All six correlate positively. But suppose the two language measures correlate especially strongly with one another. The two spatial measures also show extra similarity, as do the two reasoning measures. A general factor can summarize what all six share. Thurstone’s approach asks an additional question:
What systematic structure remains among particular groups of tests?
The answer might reveal verbal, spatial, and reasoning dimensions. Those dimensions are broader than one test but narrower than a general factor. That middle level is crucial to modern hierarchical accounts of intelligence.
Did Thurstone prove that the abilities were independent?
No. This is the point that is often lost when Thurstone is presented as the “multiple abilities” opponent of Spearman. The primary abilities themselves showed correlations. Once broad abilities are allowed to correlate, researchers can analyze those correlations too. A higher-order general dimension can then reappear. So evidence for broad abilities does not logically eliminate g. The two findings can coexist:
tests cluster into meaningful domains, and those domains are themselves positively related.
Why did factor-analysis method matter?
Factor analysis does not simply reveal a unique structure that exists independently of every analytic choice. Researchers make decisions about:
- which tests to include;
- which population to sample;
- how many factors to retain;
- how factors are rotated;
- whether factors are allowed to correlate;
- whether a hierarchy is modeled.
Thurstone’s methodological innovations made it easier to identify interpretable clusters of abilities. But if the resulting factors are forced to be statistically uncorrelated by the chosen rotation, that does not establish that the underlying human abilities are causally independent. This is an important general lesson: a modeling constraint is not automatically a psychological fact.
The rivalry became less absolute
The historical Spearman-versus-Thurstone contrast is often taught too cleanly. Thurstone’s primary abilities were not perfectly independent, and once their correlations were examined, a higher-order general factor could be recovered. Spearman, meanwhile, did not deny that particular tasks had specific demands. The disagreement did not vanish, but the empirical picture became harder to force into a binary choice. Broad abilities could be psychologically meaningful while still participating in a wider positive manifold. That convergence prepared the ground for later hierarchical models.
How did Thurstone change the one-versus-many debate?
He helped make the binary framing untenable. If only g mattered, recurring verbal, spatial, numerical, memory, reasoning, and speed-related groupings would be hard to explain. If broad abilities were completely independent, their positive correlations would be hard to explain. Later hierarchical theories effectively preserve both insights:
narrow performances → broad abilities → general covariance
In that sense, modern hierarchical psychometrics is closer to a reconciliation than a declaration that either Spearman or Thurstone simply won.
Are Thurstone’s seven abilities the modern accepted taxonomy?
No. Later work reorganized and expanded the broad-ability landscape. Cattell and Horn developed the distinction between fluid and crystallized intelligence and a wider set of broad abilities. Carroll later synthesized hundreds of factor-analytic datasets into a three-stratum hierarchy. CHC frameworks integrated parts of those traditions. Thurstone’s exact list is therefore historically foundational, not the final current map.
From rivalry to hierarchy
Thurstone’s enduring contribution was to show why a scientifically useful account of intelligence needs broad abilities as well as general covariance. His primary mental abilities helped reveal the middle level between individual tasks and g. The later finding that those broad abilities are correlated does not make them unreal. It is one of the reasons modern theories describe intelligence hierarchically.
References
- Thurstone, L. L. (1934). The vectors of mind. Psychological Review, 41, 1–32. DOI
- Thurstone, L. L. (1938). Primary Mental Abilities. University of Chicago Press. Source
- Thurstone, L. L., & Thurstone, T. G. (1941). Factorial Studies of Intelligence. University of Chicago Press.
- Carroll, J. B. (1993). Human Cognitive Abilities: A Survey of Factor-Analytic Studies. Cambridge University Press. DOI
