Wavelength generalization and preference in monochromatically reared ducklings.
Early sensory diet bends later stimulus control curves, so check rearing history before you trust a flat generalization gradient.
01Research in Context
What this study did
Researchers raised baby ducklings under only one color of light.
They then tested if the birds would peck at other colors.
Each duckling saw a range of wavelengths to map its generalization curve.
What they found
The single-color upbringing flattened the birds' generalization curves.
Ducklings raised in green light later picked shorter greens than normal.
Yet every bird differed; some curves were almost flat, others steep.
How this fits with other research
Hartley et al. (2014) extends these animal findings to children.
Their autistic preschoolers also showed flattened word-picture generalization, hinting that early input shapes stimulus control across species.
Varley et al. (1980) used pigeons and brightness instead of color, yet the same rule held: early contrast history set the gradient shape.
Hawkes et al. (1974) found peak shift in pigeons under varied reinforcement density, showing that both wavelength and reward history sculpt gradients.
Why it matters
Your learner’s early sensory world matters as much as the task you run today.
If a child was raised with limited color input, expect odd generalization.
Probe gradients early and retrain broadly to fix skewed stimulus control.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Run a quick probe of color or shape generalization with one learner and note any odd peaks or flats.
02At a glance
03Original abstract
Two experiments determined the effects of early color experience on gradients of wavelength generalization. In each experiment, one group of ducklings was raised in monochromatic (589 nanometers) sodium-vapor light and a second group, in white light. In Exp. I, ducklings pecked a key transilluminated by 589 nanometers. In a subsequent test, the group raised in white light produced steeper gradients. However, several monochromatically reared ducklings produced gradients as steep as those for the white-reared ducklings. In Exp. II, ducklings pecked a white line. In a subsequent test, using a fully illuminated key, subjects in both groups responded more often to "green" (510, 530, 550, or 570 nanometers) than to "non-green" wavelengths (490, 589, 610, or 650 nanometers). Ducklings raised in monochromatic light preferred shorter "green" wavelengths than ducklings raised in white light. This difference between the "green" preferences for the two groups accounted for most of the differences between the gradients of wavelength generalization obtained from the two groups in Exp. I after training at 589 nanometers.
Journal of the experimental analysis of behavior, 1970 · doi:10.1901/jeab.1970.13-163