Color and Cognition: A Surprisingly Deep Connection

An Accidental Discovery

In 2014, a team at the University of Sussex set out to answer a question about perception: can adults be trained to experience synesthesia — the condition where, say, reading the letter "A" automatically triggers the experience of seeing red?

They weren't trying to make anyone smarter. But that's what happened.

After nine weeks of intensive training, their participants had gained approximately 12 IQ points on a culture-fair fluid intelligence test (Cattell Scale 2, SD = 16). A control group, tested at the same intervals without training, showed no change at all. The difference was statistically significant (p = 0.021) with a large effect size (Cohen's d = 1.21).

The researchers themselves were surprised. They noted that finding any IQ improvement in healthy young adults is "notoriously difficult" and "usually limited to those at the lower IQ range" — which didn't apply to their university student sample. The IQ finding was, in their words, "peripheral" to their main goal.

But it may be the most interesting thing in the paper.

What the Training Actually Involved

The Bor et al. (2014) study trained 14 non-synesthetic adults to associate 13 letters with specific colors. But calling it "letter-color association training" undersells what participants actually did. The regime was intensive, adaptive, and overwhelmingly focused on color memory.

Every day, five days a week for nine weeks, participants completed 4–5 adaptive tasks demanding rapid, accurate recall and identification of specific colors. They practiced remembering and reproducing exact color-letter pairings, with difficulty scaling up as accuracy improved. They identified colors under increasing time pressure. And each day, they went home and read entire e-books where all 13 trained letters appeared in their assigned colors — immersing themselves in continuous color processing even outside the lab. Performance-based incentives kept motivation high: participants earned bonuses each week for improving their color recall accuracy.

The content being trained was color, start to finish. Participants spent roughly 24 hours over nine weeks intensively encoding, discriminating, recalling, and rapidly identifying colors. By the end, 9 out of 14 reported genuinely seeing colors when looking at black letters — not imagining, but perceiving them, both in the lab and in daily life. They passed the standard diagnostic tests used to identify real synesthetes.

But the synesthesia findings, fascinating as they are, leave an open question: why did IQ increase?

The Authors' Explanation — And an Alternative

The researchers suggested that the IQ gains might stem from the working memory demands of the training: "it is of course possible that the working memory aspects, and not the synesthetic features of the training regime, generated this effect."

But consider what the "working memory" in this study actually consisted of: remembering colors. Participants weren't doing generic memory exercises — they were spending every training minute encoding, holding, comparing, and retrieving color information. The working memory was entirely in service of color processing.

There's a more specific explanation worth considering: color memory may be a form of imagery training.

Think about what happens when you have to remember the exact shade of green you saw five seconds ago. The color is no longer on screen. To recall it accurately, you're forced to generate a vivid internal perceptual representation — to see the color in your mind. Do that hundreds of times, with immediate feedback on whether your mental representation was accurate, and you're essentially doing high-repetition visual imagery training with a built-in error signal.

This would also offer a natural explanation for the synesthesia results. The participants who ended up seeing colors on black letters hadn't just memorized associations — they may have trained their visual imagery system to the point where it could generate perceptual-quality color representations automatically. Under this reading, the synesthesia wasn't a quirky side effect; it was a sign of a substantially strengthened imagery capacity.

And if the training strengthened visual imagery broadly — not just for colors, but the underlying capacity for vivid mental representation — then the IQ gains would start to make sense. Mental visualization and simulation are used throughout fluid intelligence: spatial reasoning, mental rotation, planning, working through hypotheticals, holding and manipulating abstract structures. A more powerful imagery system could plausibly enhance performance across exactly the kinds of tasks measured by culture-fair IQ tests.

This interpretation is further supported by an independent line of evidence — one that rarely enters the cognitive training conversation.

The FM100 Connection

The Farnsworth-Munsell 100 Hue Test (FM100) is a pure test of color discrimination — participants arrange colored caps in order of hue. No reasoning, no memory load, no verbal ability. Just the capacity to perceive fine differences between colors.

And yet, FM100 performance correlates with nonverbal IQ more strongly than many cognitive tests correlate with each other.

In a study of autistic and typically developing children, FM100 performance correlated with nonverbal IQ (measured by WISC) at r = 0.73 — accounting for over 53% of the variance in intelligence scores (Heaton et al., 2014). In a separate study of adults, the correlation with nonverbal IQ (measured by WAIS) was essentially identical: r = 0.72 (p = 0.001; Dobkins et al., 2018).

To put those numbers in context: the correlation between the Vocabulary and Matrix Reasoning subtests of the WAIS — two core components of the same IQ battery — is typically around r = 0.50–0.60. A test that measures nothing but how well you can tell similar colors apart predicts nonverbal intelligence more strongly than some intelligence subtests predict each other.

Why would color discrimination track so tightly with intelligence? The imagery hypothesis offers a plausible link. Fine color discrimination likely relies on the same perceptual processing and internal representation systems that support vivid mental imagery. People who represent colors more precisely in their minds may also represent spatial relationships, abstract patterns, and hypothetical scenarios more precisely — exactly the capacities that drive performance on fluid intelligence measures.

Three Lines of Evidence, One Direction

Stepping back, three independent findings converge:

1. Training: Nine weeks of intensive color memory training — which we believe is fundamentally imagery training — produced approximately 12 IQ points of gain in healthy young adults.

2. Correlation: The ability to discriminate colors finely (FM100) is one of the strongest single-measure correlates of nonverbal intelligence in the literature (r ≈ 0.72–0.73).

3. Phenomenology: The same training that increased IQ also produced genuine perceptual experiences — participants seeing colors that weren't there — suggesting that their visual imagery system had been substantially strengthened in the process.

These came from different labs, different populations, and different methodologies. They weren't designed to speak to each other. But they're consistent with the same interpretation: intensive color memory training may strengthen the visual imagery system, and a stronger imagery system may in turn support better fluid reasoning.

The FM100 correlations are cross-sectional, and the imagery mechanism hasn't been directly tested in isolation. But the convergence across three independent lines of evidence is striking — and it suggests that color memory training, done rigorously and adaptively, may be a uniquely effective way to strengthen the imagery system that underlies fluid reasoning.

Training Color Memory Directly

This evidence is part of why we built MemoHue — a dedicated color memory training tool available in Relatoria's Remember module.

MemoHue takes the core mechanic from the Bor et al. study — see a color, hold it in mind, then identify it from memory — and builds it into an adaptive training system. Difficulty is controlled by making the alternative colors increasingly similar in perceptual color space (CIE LUV), so as you improve, the discriminations become finer and the internal representations you need to generate become more precise. Multiple training modes progress from single-color recall to sequential color memory, where you must hold and reproduce entire sequences of colors in order — scaling the imagery demands systematically.

Whether the imagery hypothesis holds up to further scrutiny remains to be seen. But the data is compelling enough that we think this kind of training deserves to exist — and to be freely available to anyone curious enough to try it.

References

• Synesthesia training and IQ gains: Bor, D., Rothen, N., Schwartzman, D.J., Clayton, S. & Seth, A.K. Adults Can Be Trained to Acquire Synesthetic Experiences. Scientific Reports 4, 7089 (2014). DOI: 10.1038/srep07089

• FM100 and nonverbal IQ (children): Heaton, P., Ludlow, A. & Roberson, D. Performance on the Farnsworth-Munsell 100-Hue Test Is Significantly Related to Nonverbal IQ. Investigative Ophthalmology & Visual Science (2014). DOI: 10.1167/iovs.14-16094

• FM100 and nonverbal IQ (adults): Dobkins, K.R. et al. Links between global and local shape perception, coloured backgrounds, colour discrimination, and non-verbal IQ. Vision Research (2018). DOI: 10.1016/j.visres.2018.02.004