A procedure for generating differential "sample" responding without different exteroceptive stimuli.

Eugenia Gras et al. (2003) worked with pigeons on a matching-to-sample task.

The birds first pecked under two different rules. On some trials they had to pause a long time before pecking (DRL). On other trials they had to peck quickly many times (FR).

No outside lights or sounds told the bird which rule was in force. The only cue was the pattern of its own pecks. Later, these same peck patterns were used as the "sample" in a matching game.

The pigeons learned to choose the correct comparison picture after each peck pattern.

In other words, the birds’ own response style acted like a color or shape cue. The different peck patterns had entered an equivalence class.

Tassé et al. (2013) later showed the same thing with stronger tests. They added symmetry and expansion probes and still saw that responses joined the class. Their work supports Eugenia Gras et al. (2003) with cleaner evidence.

van Schrojenstein Lantman-de Valk et al. (2006) looked like a flat contradiction. They found zero evidence that response patterns enter equivalence classes. The difference is method: M et al. used the pattern as the sample cue, while van Schrojenstein Lantman-de Valk et al. (2006) tried to make the pattern just another member after training. The tasks were not the same, so both papers can be right.

Ahlborn et al. (2008) added a boundary rule: keep the response the same for every sample inside a class. If you let the response vary too much, equivalence breaks. This tweak builds directly on Eugenia Gras et al. (2003) by showing when the trick works and when it fails.

You can create conditional cues without adding extra lights, pictures, or words. Just shape a different response topography for each set. This saves stimulus set-up time and gives you a new way to test true stimulus equivalence. Try it in intraverbal or listener training: have the child tap once vs tap five times, then use that difference to cue the next choice.