Just as humans teach their children to speak, juvenile birds learn to sing from their parents. In zebra finches, songs are the preserve of the male, and are passed from father to son. When young males are deprived of such training, their song is a poor rendition of that typical of the species. So what does that mean for generations to come if these untrained singers take on the role of tutor? For her doctoral dissertation at the City University of New York, Olga Fehér set out to find an answer.

Fehér's experiment involved having an 'isolate' male zebra finch — one reared in the absence of birdsong — raise a brood. One of the finch's progeny went on to become the main breeding male for the entire colony. Over three to four generations, each subsequent set of offspring came progressively closer to singing the species-specific song. From this finding, Fehér concludes that a culturally transmitted behaviour such as birdsong is determined in part by genetics and in part by environment.

The experiment's success relied on access to the latest sound-analysis technology. Previous studies of birdsong from untrained versus trained birds were able to merely eyeball the differences in 'syllables' as represented on a sonogram. “But that really doesn't cut it when you want to quantify subtle changes in birds' songs,” Fehér says.

To obtain more precise measurements, Fehér took advantage of a long-standing collaboration between her PhD adviser, Ofer Tchernichovski, and Partha Mitra, a theoretical physicist at Cold Spring Harbor Laboratory in New York. Mitra helped Tchernichovski to develop sound-analysis software for studying birdsong. It took into account ten different variables of sound, including frequency variation and amplitude. As a result, Fehér was able to compare different birds' songs using objective measurements (see page 564).

Other experimental methods were less sophisticated, albeit equally effective. Fehér needed a low-budget, sound-attenuating chamber in which to keep her song-naive bird colony. After searching the streets of New York City for discarded refrigerators, she found one within her building. She laid the refrigerator on its side and had it equipped with ventilation, lighting and microphones, as well as perches, swings, nest boxes and food.

Mating the isolate male was not easy. At first, none of three females placed in the cage showed any interest in him, possibly because they were accustomed to normal zebra-finch song and were not attracted to his unusual vocalizations. “We don't know much about their preferences,” Fehér says of her female finches, “but one of them eventually took to the male”.

After that, Fehér was able to let the colony evolve naturally. “Who learned from whom wasn't really controlled. What fascinated me was that, looking at the sound, it seemed as if the male offspring were imitating their father. But they obviously also had input from their siblings, especially those from previous clutches.” Such influences, Fehér believes, are what led to the gradual progression towards wild-type song.

Although the nature-versus-nurture debate was not the inspiration for the study, Fehér says, “it's an interesting example of this interaction, because it shows how strong the input is from the nature part. But it doesn't work without the nurture.”