starling image 1

Starling Talk
Care and Rehabilitation of Injured and
Orphaned Starlings

starling image 2
|| Home | Baby Starlings | Adult Diet | Pet Starlings | Health | Photos |
Links |Starlings | Sparrows | Message Board |

Learning to Sing

One of the things that is very hard for people who are raising a wild baby bird to understand is that without adult birds of its species for it to relate to, the baby will imprint on the person raising it and has no way to learn its song, the skills of survival or the ability to join others of its species in the wild. This is called human imprinting, and a bird that has been imprinted by humans will not survive in the wild. All birds that have been tested imprint. This means that a bird becomes imprinted on either birds, humans or even an object, depending on who is raising them. Research has also shown that most birds learn the song of their species at a very young age. A baby bird, like a human baby, needs to learn its language. The links below are to articles and a NPR radio program on the research into song learning in song sparrows.

PBR-All things Considered

Singing in the brain, University of Utah

Scientists Teach Sparrows to Sing Backward

Name that tune: How birds learn to recognize song

Song in European Starlings has been studied extensively in laboratory settings. It has been found that starlings require "social teaching" by other starlings for them to learn their song, as they are able to learn only a little from taped songs. Click on the links below to read some of these research articles in your Adobe Reader.

Behavioural Biology Institute of Behavioural Biology, Free University of Berlin, Haderslebener Str. 9, D-12163 Berlin, Germany The Auk 113(2):450-456, 1996

Indiana University Interspecies Interaction: A Tool for the study of Mimicry and Species-typical Birdsong Mariannne S. Engle, Muskingum College Meredith J. West, Indiana University

The following article was presented at a recent International Wildlife Rehabilitation Council conference in Portland, Oregon. My appreciation and thanks to Ms Dolinsky for permission to use it on this page.

Missing a Crucial Step?

Department of Biology,

University of Miami

Abstract: The process of learning to sing in passerines is not an instinctual behavior, but is instead learned from their parents. Songbirds learn to sing in their first 60 days. In a rehabilitation setting, songbirds are missing this crucial part of their development. Without the proper song development, songbirds are unlikely to gain a territory, or to attract a mate.


Songbirds include those birds from the order Passeriformes, commonly called passerines. There are approximately 5,400 different species of passerines. The suborder Oscine includes those birds that we tend to think of as your typical songbird, those birds that have complex songs as adults. In the majority of Oscines, the male is the one with the complex singing ability, and there is much diversity among males of different species in the length and complexity of their songs.

Song needs to be developed, just like any other characteristic of songbird behavior. Singing the correct song for that species is not an instinctual process. The need to sing is instinctual, however what song to sing has been found to be learned. When and where young songbirds learn their songs has long been studied. Just as songbirds differ in their songs, they also differ in their song learning process, though some basic principles seem to be the same. Many species learn their song only during the first few months, though a few can learn songs their entire lives. Also differing between species is from who the songs are learned, and how much of the songs are accurately imitated (Liu & Kroodsma 1999).

Some species may exactly copy the songs they are exposed to. On the other hand, many will not exactly imitate the song or songs that they have been exposed to. Instead they will take bits and pieces of it and invent their own song, though usually similar in construction to that which the species normally sings.

There are several different types of songs that I will be referring to. Learned songs are just that, songs that have been learned from some source. There are also isolate songs. When birds are raised without hearing an adult of the same species, they tend to develop this type of song. Isolate songs are typically much simpler, and are composed of longer, similarly pitched notes, than are the natural, learned songs of the species.


The Basic Features of Song Learning –-

A Case Study I am going to introduce the learning process giving a case study and will then come back later to go over specific aspects of the learning process. The majority of research on this topic has been on many of the native sparrow species. For this case study we will be examining an experiment carried out involving song sparrows.

Nine males were collected from the wild from four different broods when they were around 4 to 6 days old. They were raised by people until they were weaned at 33 to 35 days old and then were placed into individual wire-mesh cages. Live tutors were used in this experiment. Four wild-caught adult male song sparrows were used as the tutors. The tutors were placed into flight aviaries and the subject’s cages were placed adjacent to the aviaries, so that they could have visual contact with one of the tutors, while still being able to hear the other three singing. Also, the subjects were rotated so that their visual contact with one tutor varied between the tutors. The subjects were there when they were 33 to 94 days old, with the assumption that their sensitive phase lay some time in this time period.

When the subjects started singing the next spring, their repertoires were analyzed. The average was seven song types per subject. This does correspond with birds raised exclusively in the wild by their parents. Eight of the nine subjects learned songs from two or more of the tutors and later imitated these songs (Nordby et al. 2000).

In similar experiments in which some birds are exposed to tutors and some are raised in isolation, those birds that were exposed to tutors almost always tend to imitate pieces and even complete songs. Those birds raised in isolation always sang isolate songs. These isolate songs were similar to each other and consist of much more structurally simpler songs than males raised with normal song exposure (Nowicki et al. 1999).

Sensitive Learning Period

There seems to be a specific time period in which songbirds learn their songs. This “sensitive phase” is “when an individual hears and is thought to acquire song models” (Peters et al. 1992). Most species learn their songs as juveniles. This learning period lasts from around ten to sixty days of age. This learning period corresponds very closely with the age at which baby birds are in a rehabilitators care. The songs that they are exposed to during this time period are what they will learn from and start singing once they mature. Songbirds do not start to sing until the next spring, once they are around three hundred days old. Once they start to sing, they only imitate songs that they heard during their early sensitive phase. Just a few passerines, such as the European Starling, “are capable of learning songs later in life” (Liu & Kroodsma 1999).

There is some controversy as to whether live tutoring leads to a longer sensitive phase than does tape tutoring. However, recent experiments show that there is little to support this view. (Nelson 1998; Catchpole & Slater 1995).

Live Tutors versus Tape Tutors

Song learning has been taught successfully in the lab in a variety of ways. The two most commonly used teaching practices are by using a tape tutor and the use of live tutors. Tape tutoring is playing a recording of a song, or multiple songs over and over. Live tutors are adult males of the same species kept either in the same cage as the babies, or in adjacent cages so that the young can both interact with and hear the adults singing (Catchpole & Slater 1995).

Many species can learn songs equally well from either taped songs or live tutors. Some species, however, have had limited, if not zero success learning songs from a tape and have only been recorded in lab learning from live tutors. In addition, the age at which learning occurs may differ from live to tape tutors. This difference in learning success, however, may have to do with how often the birds are exposed to the tape versus the live tutor. If they are with the live tutor all day long and he is singing all day long, there will be much more exposure than just hearing a taped song a few times during the day (Nelson 1998).

Studies performed on hand reared chipping sparrows and field sparrows have shown no learning from tape-recorded songs. Instead their songs were highly abnormal and were similar to babies raised without the benefit of hearing any song whatsoever. When exposed to live tutors, however, both species have been shown to learn to properly sing (Liu & Kroodsma 1999).

Learning From Limited Exposure to Tutors

The question now arises as to how often babies need to be exposed to songs in order to learn them. Most experiments “expose test subjects to hundreds or even thousands of repetitions of tape-recorded song models over the course of training” (Peters et al. 1992). There have been a few studies performed in which babies were exposed to much less tutoring. One study done on song sparrows showed that “song sparrows are capable of learning from 30 repetitions of a song type heard in a single 5-minute bout” (Peters et al. 1992). The experiment was performed on nine hand-raised song sparrows. The subjects were raised out of auditory contact with other birds until they reached 50days of age. This experiment exposed song sparrows to taped songs varying from over 1000 presentations of the song for a four week time period, to 30 presentations of a song over one five minute time period. Both groups learned their songs equally well. Also, some of the most successful learning experiments only played the songs to the birds twice a day, once in the morning, and once in the evening. Of course, one must keep in mind that this experiment was only performed on song sparrows, and the results may differ dramatically in other songbird species.

How Birds Learn

Since we have now gone over, what and when birds learn, next we will go over how birds learn. Birds exposed to songs of a different species do not start imitating that song. Instead they seem to know to only imitate songs coming from conspecifics. They have the ability to discriminate between songs of their own species and those of other species. Interestingly enough, nestlings and fledglings even respond differently when exposed to their song versus a song from another species. In one experiment, “fledgling white-crowned sparrows called more vigorously in response to playbacks of songs containing conspecific phrases than to playbacks of songs containing heterospecific phrases” (Soha & Marler 2001). Somehow, babies are able to discriminate between their song and others, even prior to the sensitive phase of learning.

Research seems to suggest that there are specific acoustic cues involved in the learning process. These cues seem to be in the length, tone and pitch of the introductory whistle of a song. In fact, one group of researchers was able to get white-crowned sparrows to successfully imitate heterospecific songs, by just adding their introductory whistle. Certain conspecific phrases are used as cues to discriminate between conspecific and heterospecific songs (Soha & Marler 2000; Soha & Marler 2001).


The significance of song learning may have several implications for rehabilitators. Songbird vocalizations are important socially in a variety of ways. Birds that have not learned to sing will most likely never be able to successfully attract a mate. In addition, singing is essential when gaining a territory. Those birds that are unable to correctly sing will not be able to out-compete others for a territory. Without a mate or a territory, we must ask ourselves how normal of a life these birds will lead. Sure, they are alive out in the wild thanks to us, but as rehabbers isn’t it is our responsibility to teach our patients to lead as normal a life as possible once they are released and are on their own?

So, as rehabbers, what are some simple and affordable solutions? Several methods may be used to help rehabilitated babies learn to sing. Tape recordings of most songbirds exist and are easily found for sale. These may be purchased and then played to babies, especially in the fledgling stage. My favorite method is using the ‘Audubon stuffed animals’, those birds that play back a real recording of a bird song. It is very simple to keep a stuffed animal of that species on top of a cage, and then play the stuffed animals every time you are feeding that cageful of babies, or every time you walk by. In addition to being practical, these birds are also cute and visually pleasing for any visiting members of the public, and they are a wonderful way of distracting those annoying kids that want to hold one of the cute baby birds. Give them a stuffed animal to hold instead! Also, most bird songs are easily found online at sites such as the Cornell Lab of Ornithology site, or at Enature. You can also buy a portable, handheld birdsong player called the BirdSong Identifier, made by For the Birds. These come with cards that you insert. Each card has 10 songs on it. One could easily play these with one hand while feeding babies with the other. The full set of songs and player is a bit pricey and costs around $150 with the player alone costing approximately $35. Playing recorded songs will easily work for smaller rehabbers, however for larger rehabilitation centers I’m not sure how practical this would be. Also, tape recordings have not been proven capable of teaching all songbird species.

A more effective solution seems to be housing fledglings with adult males of the same species. Baby songbirds raised in this way have a much higher success in learning to sing. They can also be housed in adjacent cages, so that the babies can see the adults, and of course house them adjacent so that the babies can at the very least hear the adults. The highest learning results have been in those situations in which the baby birds have been housed in the same cage as the adults. This subject definitely makes the argument for keeping non-releasable songbirds in order to use them as tutors.

While there is still much controversy surrounding the different aspects of teaching songbirds to sing, some aspects can be generalized. The process of singing definitely appears to be a learned one and is not instinctual for most songbirds. Also, there is a specific period in which songbirds are open to learning new songs. This appears to lie within the first sixty days, though there is a lot of variation between species. Rehabilitators have baby birds in their care until they are around one and a half to two months old. This almost exactly corresponds with this sensitive learning phase. Rehabilitators are currently sending songbirds out into the wild without teaching them the crucial behavior of singing. Without being able to attract a mate or gain a territory, these birds will live a life of isolation. Also, if none of our babies are contributing to the future gene pool or population growth of the many declining species, then raising songbird orphans will have no conservation impact whatsoever.


Thank you so much to the University of Miami for their support in this paper. Specifically, I would like to thank Dr. William Searcy for his help in finding materials and pointing me in the right direction in finding information on this topic. I am also very grateful to Dr. Ted Fleming for reviewing this paper and for his helpful comments.


Catchpole, C. K. & Slater, P. B. J. 1995. How song develops. In: Bird Song, pp. 45-69. Great Britain: Cambridge University Press.

Liu, W. & Kroodsma, D. E. 1999. Song development by chipping sparrows and field sparrows. Animal Behaviour, 57, 1275-1286.

Nelson, D. A. 1998. External validity and experiment design: the sensitive phase for song learning. Animal Behaviour, 56, 487-491.

Nordby, J. C., Campbell, S. E., Burt, J. M. & Beecher, M. D. 2000. Social Influences during song development in the song sparrow: a laboratory experiment simulating field conditions. Animal Behaviour, 59, 1187-1197.

Nowicki, S., Peters, S., Searcy, W. A. & Clayton, C. 1999. The development of within-song type variation in song sparrows. Animal Behaviour, 57, 1257-1264.

Peters, S., Marler, P. & Nowicki, S. 1992. Song sparrows learn from limited exposure to song models. Condor, 94, 1016-1019.

Soha, J. A. & Marler, P. 2000. A species-specific acoustic cue for selective song learning in the white-crowned sparrow. Animal Behaviour, 60, 297-306.

Soha, J. A. & Marler, P. 2001. Cues for Early Discrimination of conspecific song in the white-crowned sparrow (Zonotrichia leucophrys). Ethology, 107, 813-826


To Starling Talk homepage.

Copyright © 1999-2011 Jackie Collins - All rights reserved.