Humans exhibit associative symmetry in the absence of backward training and stimulus overlap

Navarro et al. (2025) asked a simple question. Does symmetry really come from the way we train it, or is it a true human skill?

They removed two usual helpers: backward training and stimulus overlap. Adults learned A-B and B-C relations only. Then they tested if A-C and C-A emerged without ever seeing those pairs.

Two small experiments used the new setup. Both aimed to see if bidirectional links still formed.

Symmetry showed up strong and fast. Adults could match C back to A even though they had never been trained that way.

Because the usual crutches were gone, the effect looks real, not a trick of the procedure.

Campos et al. (2011) saw symmetry in pigeons, but the birds failed on transitivity. Humans in the new study passed both, showing a species gap.

Older work by C et al. (1990, 1995) flipped baseline contingencies. Symmetry broke while transitivity held. Navarro’s cleaner method keeps symmetry intact, hinting past breaks were artifact, not law.

Austin et al. (2015) proved equivalence works for kids with autism. The new paper shows the same core process holds for neurotypical adults once you strip the clutter.

Lionello‐DeNolf (2021) reviewed 16 animal studies and found symmetry only 30% of the time. The fresh human data strengthen the claim that robust symmetry is mainly a human feat.

You can now trust symmetry as a genuine learning outcome, not a by-product of your training order. When you write lesson plans, skip extra backward trials and save time. Probe bidirectional relations early; if they emerge, move on. This keeps therapy efficient for any learner who can handle conditional discrimination.