Add-1·Add-3 pupil experiment

The Add-1/Add-3 pupil experiment made an invisible process—mental effort—physically observable. Participants transformed strings of digits under a metronome’s relentless rhythm while an infrared camera recorded their pupils. The resulting pattern linked subjective strain, task difficulty, pupil size, and heart rate, establishing a vivid physiological index of cognitive work and helping illuminate the division between effortless intuition and the effortful operations of System 2.

Task Design and Experimental Demands

The basic Add-1 task required participants to prepare strings of four different digits, read each string aloud, wait for two beats, and then report a new string in which every digit had been increased by one. A card showing 5294 therefore required the response 6305. The participant had to preserve the rhythm while holding the original digits in short-term memory, waiting through the pause, and transforming each digit in sequence. The metronome was not decorative: its one-beat-per-second pace fixed the timing of the mental operations and made it possible to know what the participant was trying to remember or calculate at each moment.

The experiment also included Add-3, in which each digit had to be increased by three rather than one. Few people could manage more than four digits in Add-1, and Add-3 provided a substantially harder challenge. The task therefore produced controlled changes in cognitive load rather than relying on vague reports of whether a participant felt tired. Longer strings increased the amount of information to be held, while transformation added another operation to the memory burden.

Measuring Effort Through the Pupil

The experiment grew from the observation of Eckhard Hess that pupils dilate during demanding mental multiplication and dilate more for difficult problems than for easy ones. Daniel Kahneman and Jackson Beatty adapted this insight into an apparatus resembling an optician’s examination room. A participant rested her head on a chin-and-forehead support, stared at a camera, listened to prerecorded material, and answered on the beats of a metronome. Each beat triggered an infrared flash and a photograph. After each session, the researchers projected the pupil images and measured them with a ruler.

The method was especially effective because the researchers controlled the contents and timing of the participant’s mental activity. They recorded digit strings on metronome beats and instructed the participant to repeat or transform the digits one by one while maintaining the rhythm. The pupil changed from second to second in correspondence with the changing demands of the task. Its response formed an inverted V: effort accumulated as each additional digit arrived, reached its highest point while the participant rushed to produce the transformed string during and immediately after the pause, and then declined as the contents of short-term memory were unloaded.

Findings and Physiological Limits

Pupil size tracked both the length and the kind of operation required. Longer strings produced larger dilations, and transforming the digits demanded more effort than merely retaining them. The largest pupil size coincided with the point of maximum effort, while the measurements corresponded closely to participants’ reported experience of strain. Add-1 with four digits caused a larger dilation than holding seven digits for immediate recall, showing that the cognitive operation performed on information could matter more than the number of items held.

Add-3 marked the upper boundary of the work that the researchers observed. During its first five seconds, the pupil expanded by approximately 50 percent of its original area, while heart rate rose by about seven beats per minute. The response represented an extreme level of effort: when still more was demanded, participants gave up. If they were exposed to more digits than they could remember, their pupils no longer dilated and could actually shrink. The result distinguished intense but organized mental work from overload; beyond a certain point, additional demands did not produce more effortful performance but the collapse of the effort itself.

Broader Significance

The experiment translated the abstract language of cognitive effort into a visible temporal record. It showed that mental work is not a uniform state but a sequence with an onset, accumulation, peak, and release. Because the researchers knew precisely what was on the participant’s mind at each beat, they could connect a bodily signal to identifiable stages of remembering, transforming, and responding rather than treating pupil dilation as an undifferentiated sign of arousal.

The findings also supplied a concrete image for the book’s account of System 2. System 2 performs operations that require attention, working memory, and self-control, but its effort is limited and it is reluctant to spend more energy than necessary. Add-1 and Add-3 reveal that limitation from the outside: the pupil enlarges as the task becomes demanding, reaches a sharply defined maximum, and ceases to enlarge when the task exceeds the participant’s capacity. The experiment thus stands as a physiological counterpart to the broader study of heuristics and biases, where the mind often avoids such effort by allowing System 1 to generate quick, intuitive answers instead.