Investigations of a mirror-image transfer effect in pigeons.

Stretch et al. (1966) trained pigeons to peck when they saw a line tilted one way.

Later the birds saw lines at many tilts. The team drew a graph of how often the birds pecked at each new tilt.

The graph had two tall peaks. One peak sat at the trained tilt. The other peak sat at the exact mirror tilt.

The birds acted as if the mirror image was the same as the training image.

Mello (1966) saw the same mirror effect when pigeons used only one eye. That paper is a direct replication; it shows the effect is not just a quirky result.

Schroeder et al. (1969) later showed that waiting a day before testing flattens these peaks unless you give a quick warm-up. Their work extends the 1966 finding by adding a time rule.

Campos et al. (2011) tried to build full stimulus equivalence in pigeons but only got symmetry. Mirror-image pecking looks like symmetry, yet full equivalence fails—so the bird brain may stop at mirrored forms.

When you teach a child to pick the letter b, watch for pecks at the letter d. Mirror generalization is built into visual learning. Test both forms early so you can shape clear discrimination before errors pile up.