This is Scientific American 60 Second Science.

I'm Christopher Intagiyata.

In the most recent podcast, we discussed how baby bats learn their calls

from all the other bats in their crowded colonies. And we mentioned in passing that... their work.

At about 25 days, the father starts singing in many cases directly to the juvenile. David Met, a geneticist at the University of California, San Francisco.

That sort of is the onset of what's called the sensory phase of learning where they incorporate

information from their environment.

What Mets and his team wanted to know was how much of a baby bird's future musicality

is influenced by that tutoring, an environmental factor, and how much is written in their genes.

So they studied Bengalese finches, which sing like this.

The tempo of that song appears to vary according to a finch's genetics.

So they tried training baby finches with different genetic tendencies, fast, medium, or slow singing,

on a synthetic finch song made from a library of different types of song syllables.

So these are sort of, it'd be tonal downward sweep, so, you know, or sort of broadband noisy ones like shh.

But when baby finches with different genetic backgrounds were trained on the resulting tune,

the training didn't stick.

Instead, the greatest predictor of their singing tempo was the way their fathers

sang, which they'd never heard. So their genes seemed to be in charge. But then Metz flipped the experiment,

exposing genetically similar birds to actual live birds that sang

fast, medium, or slow, and that live training appears to have been compelling enough to override

the influence of the bird's genetics,

so that genetically identical chicks sang tunes fast, medium, or slow,

...