I recall hearing about newborn birdsongs being learned "in utero" (not sure if quite the right term but lets go with that). In that case the channel for transmission was sound. It was apparently used as a shibboleth against brood parasite egg replacement. If the baby didn't sing the song that was being sung to it then the baby got abandoned by presumably disappointed parents. I suppose it could also be a 'health test' of sorts since sufficiently deformed or disabled offspring would also fail.

Parental teaching and learning is a spectrum and not a binary. We've found with relocating deer (to similar but not identical environments) doing worse until learning occurs over a few generations and they catch up. Animals may not be as intelligent as us but their ability to learn and adapt should not be underestimated.