VOCAL MIMICRY in COMPANION PARROTS
Species Differences In Vocal Learning By Parrots Revealed By Citizen Scientists
by GrrlScientist / Dec 7, 2022
“A very interesting study was just published that finds many parrots use words in appropriate contexts. This study, based on an analysis of almost 900 companion parrots as reported by a flock of citizen scientists, provides the largest comparative analysis to date of companion parrot vocal repertoires.
“I spend a lot of time thinking about male and female vocalizations in songbirds, so I first approached parrots from that point of view”, study co-author, Lauryn Benedict, a Professor and Associate Director of UNC’s School of Biological Sciences at the University of Northern Colorado (UNC) said in email. “I was interested in a group where the general wisdom was that males and females were equally good vocal learners”, Professor Benedict explained. Currently, the general wisdom is that male songbirds have better vocal abilities than females because they use their song to attract a mate and to defend their territories. In parrots, on the other hand, both sexes may be equally chatty. “I wondered if that was really true [for parrots] and thought that a comparative study across species would be useful.”
In the wild, parrots and their closest evolutionary relatives, songbirds, learn their vocalizations from their families, neighbors and communities. But it’s very difficult to study parrots in the wild so Professor Benedict thought it would be impossible to get enough standardized data for so many different species. She contacted Christine Dahlin, who researches parrots in the wild, for advice. “[As] a field biologist, I am well aware of how hard it is to study wild parrots and acquire good data”, study leader Christine Dahlin, an Associate Professor of Biology at the University of Pittsburgh, Johnstown (UPJ), told me in email. “Parrots live in tropical places where it is often difficult to get access to them, they fly vast distances, they blend in with trees, and it is super hard to band them or otherwise identify individuals. So the vast amount of data we have on songbird communication is missing for parrots.”
But they both recognized that many thousands of captive parrots and their human companions might be willing to provide the research community access to a huge, untapped, source of information about basic parrot vocalizations. “We realized that a community science approach could get us comparable data on hundreds of birds of many species”, Professor Benedict continued in email. “Plus, we could collect all that data during a pandemic – no field work or social gatherings required. All of that seemed very appealing.” Because companion parrots are mainly exposed to human sounds — speech and other noises that fill our homes — they end up mimicking those sounds.
This availability of learned vocalizations produced by parrots led Professor Dahlin and Professor Benedict to begin asking fundamental questions about their vocal abilities — questions that many of us ask, or opinions that we accept without question. These basic questions included: how large are the vocabularies of different parrot species? Do male parrots ‘talk’ more than females? Do companion parrots continue to learn new words or phrases as they get older? Thanks to the reach of the internet and the public’s growing willingness to participate in online surveys to benefit scientific research (especially during the COVID-19 pandemic lockdown), Professor Dahlin and Professor Benedict and their collaborators created an online Google Docs questionnaire and invited parrot people to answer a few questions about their companion parrots’ vocal abilities.
(Although the online survey is still actively seeking and recording responses, this particular study was based on survey responses collected between 5 October 2020 and 18 April 2021.) Responses flooded in from hundreds of citizen scientists about the vocal abilities of their citizen scientist parrots to create the largest database on parrot vocalizations yet amassed. This survey collected standardized data for nearly 900 companion parrots of 73 species — a massive dataset that would have been impossible to assemble from wild parrots. Reported vocalizations included both words and phrases, whistled tunes and other sounds learned from a human, and also asked whether these sounds are produced in appropriate contexts. The least surprising aspect of these collected data answers the question: does species matter for vocal mimicry in parrots?
“As it turns out, Polly’s species might have a strong impact on what she says”, Professor Benedict stated. According to these data, a parrot’s species does matter when it comes to mimicking sounds — and African grey parrots are the champion blabbermouths. “Some species are much better mimics than others”, Professor Benedict pointed out. “African grey parrots, long understood to be the best at learning human sounds, were found to have the largest repertoires, averaging about 60 human words.
Cockatoos, Amazons, and macaws also were excellent mimics, with average repertoires of 20-30 words.” Could the vocal abilities of parrots reflect the amount of time and effort that people invest in teaching their companion parrots to mimic human speech? For example, people may be biased to think that grey parrots are much better mimics than, say, cockatoos, and thus, invest more time working with them. “Very likely”, Professor Benedict replied in email. “We expect that all birds living with humans have the opportunity to learn human words and sounds, but some might get more ‘teaching’ than others. This will certainly be a bias in our data set.”
Perhaps not surprising, Professor Dahlin and Professor Benedict and their collaborators found extreme variability in vocal mimicry within species, which probably reflects extreme variability in training provided as well as a parrot’s vocal abilities — and the parrot’s motivation to mimic sounds. “Survey-takers reported on Grey parrots with anywhere from 0 to 500 words”, Professor Benedict told me in email. “Much of that variability is likely due to human social interactions with the birds, along with the birds’ innate learning tendencies and biases.” Although most parrot species learned more spoken phrases than sounds, a few, such as cockatiels and Fischer’s lovebirds, seemed more adept at mimicking sounds (doorbells, whistles or digital or electronic sounds) rather than spoken phrases.
Reading this study, I was most surprised by the (relatively) poor showing of budgerigars in these data, which, according to my experiences, are outstanding mimics of human speech. (Disco the budgerigar stands out in my mind here. See more on his dedicated YouTube channel.) Additionally, as I was researching this piece, I stumbled across this interesting titbit on Google: the champion talking parrot was a male budgerigar named Puck, who currently holds the world’s record for the largest vocabulary of any bird with an incredible 1,728 words. (Just to put this accomplishment into perspective, a human needs to know and appropriately use somewhere between 300-600 words to travel, whereas at least 1,000 words are necessary for a conversation.)
Sex is not as important to a parrot’s vocal abilities as many people believe. For example, I have often been told by parrot breeders and hand-feeders (in several countries) either that female parrots don’t mimic human speech at all or that female parrots are poor mimics — but according to this study’s findings, this is a fallacy in desperate need of correction. This erroneous idea is further reinforced by a widespread reporting bias where chatty parrots of unknown sex are assumed to be male (74%) by their human companions.
Day 4 of #Parrot tweets. Parrots can learn new vocalizations throughout life. . . but 5 year-old grey parrots say as many words as 50 year-olds. Repertoires turn over. Words get added and dropped. Among wild birds this can foster social integration. https://t.co/WYMoDHRVK3 https://t.co/fR7ui19Ptb pic.twitter.com/8SetSxLfg7
— Lauryn Benedict (@LaurynBenedict) December 8, 2022
In fact, survey participants verified this idea: the researchers concluded that people who live with chatty parrots of uncertain sex overwhelmingly, and often mistakenly, assume those birds are male without actually checking this using a fool-proof method to identify sex (DNA methods, surgical sexing or sexual dichromatism). And yet, this idea that male parrots are better mimics than females really is true for a very few parrot species, such as budgerigars, where males have larger vocal repertoires, Pacific parrotlets, where only males were reported to “talk” (the small number of Pacific parrotlets in this survey may be misleading), and yellow-headed amazon parrots, where females were reported to be the chattier sex.
The study also indicated that age could be a factor in vocal mimicry, with juveniles showing the greatest expansion of their repertoires until reaching sexual maturity, when their vocal repertories reached a plateau. According to the data analyses, fifty-year old parrots did not have larger repertories than five-year-olds, although older parrots were sometimes reported to learn new words when they move in with a new family. (This was the case for two of my three grey parrots — sisters — whose spoken German was better than mine until moving in with me as adults and then learning English.) Whether parrots understand spoken human language is a bit more difficult to quantify but it becomes somewhat more convincing when listening to a parrot. Basically, is the parrot using words or phrases appropriately and in context?
In this survey, human caregivers report that 89% of their parrots spontaneously use human words or phrases in appropriate contexts, and most of them do so frequently (b). “I think our data show that they can use words and phrases in context, but that doesn’t necessarily imply understanding”, Professor Benedict cautioned me in email. Additionally (and more subtly), the survey participants report that their companion parrots demonstrate their vocal abilities by manipulating learned language to invent new words or to reorganise phrases (a). “They use [language] in context and some of them recognize words as building blocks because they can move them around in phrases”, Professor Benedict told me in email. “At the same time, we can’t conclude true understanding from our data.” We’ve long known that parrots rely upon vocal mimicry to establish and maintain their cultures and social connections, and to communicate with their families and neighbors — just like people do.
Are there parallels in how birds, machines, and humans learn to recognize bird vocalizations? Maybe. Teaching @MerlinBirdID to recognize birds by their sounds has been challenging for a variety of reasons. pic.twitter.com/0L2ccNEyQs
— Eliot Miller (@EliotITMiller) July 7, 2022
This is in stark contrast with most animals, which cannot learn to imitate sounds at all. Although a small number of evolutionarily distant mammals, such as dolphins and bats, do have this capacity, our closest relatives, the non-human primates, don’t show a similar ability to learn new sounds. But even amongst these more evolutionarily distant animals, birds — especially parrots and songbirds — excel at vocal communication. But parrots’ vocal and cognitive abilities are overlooked by most of the scientific community. As such, parrots are important (and mostly unexplored) research subjects for providing greater understanding of the physiological, neurobiological, and evolutionary underpinnings of acoustic communication. “This research highlights just how much parrots still have to teach us”, Professor Dahlin said.
But this study focuses solely on companion parrots that live in a tremendous variety of environments and who are provided vastly different levels of training by their human caregivers rather than being studied under carefully designed lab conditions. But the research team writes in their study: “Although our community science data lack the precision of intensive lab and field studies, they provide a valuable comparative resource for animal communication researchers. Future work could pair the data reported here with information on parrot neuroanatomy and ontogeny to develop this group as a model for vocal production learning, a critical component of spoken language.”
This belated realization of the value of crowd-sourced data on companion parrots further serves to highlight the growing peril that parrots are faced with today. “Approximately 30% of parrot species in the wild are declining to the point of being threatened, endangered or critically endangered, primarily from poaching and habitat loss”, Professor Dahlin pointed out. “Without conservation of remaining populations, we risk losing the opportunity to understand the evolution of complex communication in these amazing animals.”
Source: Lauryn Benedict, Alexandra Charles, Amirah Brockington and Christine R Dahlin (2022). A survey of vocal mimicry in companion parrots, Scientific Reports 12:20271 | doi:10.1038/s41598-022-24335-x
It turns out, Puerto Rican Parrots (#iguacas) have dialects. Populations in different parts of the island have different calls. Wild parrots and captive parrots also produce different calls. How did this happen? Read all about it in our new publication: https://t.co/blnupkCZLa pic.twitter.com/dA6bkBBUg0
— Tanya Martínez (@iguacachick) July 18, 2020
WILD PARROTS LEARNING to TALK
DOG TELEPATHY RESEARCH