What questions are you dealing with in this project?
Do other animals have language? So far, we do not have evidence to support the idea that other animals have a communication system as complex as language. While we know that animals can communicate important pieces of information to each other (e.g., a flying predator is approaching!) we don't know if they can have a conversation referring to something that is not currently happening (e.g., "Jim always steals food from me so I have been avoiding him"). This kind of "conversation" would be difficult to detect in part because of a lack of clear correlation between the information and the situation, but the other critical part is that we know very little about the structure of non-human animal vocalizations.
Can we discover the basic units that underlie another species' vocalizations? Most of the species that have been studied have fairly simple repetitive vocalizations. Other animals exists that, like humans, produce novel combinations of sounds each time they vocalize. While we humans say some things over and over again (e.g., "Grüß Gott!", or "Hello, how are you?") many of the sentences we say (e.g., this sentence) never get repeated in our lifetime. Budgerigars ("Wellensittiche" in German) are a small parrot species that, like humans, produce novel utterances each time they vocalize. Our goal is to find the basic units in this species' vocalizations and study how the birds make use of these units.
What results do you expect?
Our specific aims are to:
- Record budgerigars in a naturalistic situation when they are all vocalizing at the same time (kind of like at a party for humans) and then extract the individual vocalizations that each bird is making
- Assess whether these vocalizations are used in a dialogue between individuals like a human conversation. Does one bird stop and a neighbouring bird start?
- Automatically extract potential units from the vocalizations and compare them to each other to find a budgerigar "alphabet" and then test budgerigars using behavioural tests to see which units they consider the "same" and "different"
- Assess whether there are meaningful combinations of units that the budgerigars appear to repeat
On the whole, our goal is to form a foundation for understanding whether budgerigars have a communicate system similar to language that then can also be applied to other species with complex vocalizations.
What assumptions is the project based on and what expectations do you have?
We know that budgerigars use their complex vocalizations in courtship, and that they also vocalize together in a group. The same can be said for human language: we use it in courtship, and many of us are talking at once when we're at a large party. But what we don't know is whether the budgerigars are simply showing off their ability to make interesting sounds, or whether the combination of sounds they make have referential meaning the way language does. Probably the answer lies somewhere in between, and our goal is to take the first steps towards finding an answer.
What use cases and benefits do you see?
Our findings will help us determine whether humans are the only species on the planet with language or whether language is more common than we thought. This can help us understand how language evolved, and in the long run, our results could allow us to communicate with other species.
If you had the time and resources: In which direction would you like to develop this line of research?
There are 2 main long-term aims:
- once we understand the basic vocal units of budgerigar vocalizations we can build a database of video and audio recordings that would allow us to analyze the meaning, if any, of these vocalizations
- apply our protocols to other species that produce complex vocalizations (e.g., marine mammals, some songbirds and other parrots, bats, elephants etc) to understand whether and which other animals have communication systems that are similar to language.
How did you come up with the research idea and how did the actual project come about?
In previous research, my Marie Curie Postdoc, Dan Mann, discovered that budgerigars appear to have vocal units that are similar to consonants and vowels. However, he only was able to get recordings of single birds that sang into a microphone that he pointed at them (it was very cute but not very naturalistic!). We also do not yet know how the birds perceive subtypes of these units. This made it clear to us that budgerigars were an ideal study subject for determining whether another species may have something akin to language.
Is there anything that particularly encouraged the exchange and made the application possible?
The advancement of machine learning is key in the ability to carry out this project. To date, complex vocalizations in other species have typically not been studied very deeply, because it is very difficult to know where to begin. Most of the interesting findings about meaning in animal vocalizations come from stereotyped vocalizations (e.g., alarm calls indicating a specific kind of predator). By combining our efforts with mathematicians focused on acoustic signal processing, we have the perfect interdisciplinary team to finally study a complex vocalization in a naturalistic situation.
Which disciplines are involved and what role does interdisciplinarity play, where does your project fit into cognitive science research?
On the one side we have mathematicians that are at the cutting edge of algorithms currently used for processing acoustic signals including more traditional models and machine learning models. On the other side we have biologists, cognitive scientists, and linguists who have been tackling questions of the evolution of language. Without the combination of these fields it would impossible to conduct the current project.
What methods will you use?
We will use machine learning methods to identify the location of individual budgerigars in an aviary, and other machine learning/signal processing algorithms to extract the individual audio streams that come from each location in the aviary as recorded by an array of microphones (the delay in arrival of sound at the different microphones can be used as a clue to the spatial location of sound similar to how having 2 ears allows us to pinpoint the origin of a sound). Afterwards, we will use phonetic and machine learning methods to isolate units in individual budgerigar vocalizations, and behavioural testing methods to effectively ask the budgerigars how they perceive their own vocalizations.
What insights into cognitive processes in humans do you hope to gain from your research? (How) Does understanding animal communication help us to understand human communication?
Our project will allow us to draw parallels between human language and the communication system of budgerigars. This kind of comparative cross-species research is one way to understand the origins of human behaviours. Because we humans are just one species on the planet, and because our capacity for language is based in our biology (e.g., isolated humans form new full-blown languages within 2 generations) by comparing our capacities with those of other species we can gain insight into why they may have arisen. We already know of several important parallels that budgerigars have with humans in terms of their communication system that are uncommon in the animal kingdom. Budgerigars learn their vocalizations from adult members of their species just like human children learn language. Separated groups of budgerigars have cultural differences in their vocalizations known as "dialects", much like how humans have different languages in different regions. Budgerigars can adapt their vocalizations when entering a new group of individuals, learning to vocalize in the local dialect. Budgerigars have complex social systems without simple hierarchies. They form long-term monogamous pair bonds, but they will not necessarily stay together for life. Especially if the group composition changes, the birds will sometimes move to a new partner, and drama often comes along with this. Are these birds having conversations the way we are? Why would a creature so distantly related to humans display so many parallels in communication? If we find parallels in budgerigar communication to human language, this would suggest that 1) there are likely to be critical parallels in the ecological niches that humans and budgerigars fill that drive the evolution of complex communication systems and 2) the parallels may outline some of the key cognitive constraints that all species requiring a complex communication system face. This could allow us to find complex communication systems in other modalities (e.g., visual or tactile). Ultimately, this will allow us to gain insight into under what conditions a species develops language, and what fundamental parallels all languages may have.
For more info about this approach and a review of the work we conducted that motivated this project, see: https://pubmed.ncbi.nlm.nih.gov/36574158/
What do you think made your project so competitive, how did you convince the WWTF? What tips do you have for young researchers who want to assert themselves in the field of interdisciplinary research?
I think in part we were lucky that a call happened to come out that perfectly fit a project idea we had intended to write a grant for anyway. So we already had everything in place to conduct the work, all we really needed was the funding to hire people to dedicate their time to the project. Because the WWTF mainly wants to fund personnel in interdisciplinary projects, and because we were able to present a well-thought-through and realistic plan that fit perfectly with the call, I think that may have given us an edge in this competition. My advice to young researchers would be to look for positions in different kinds of departments to gain insight how approaching the same problem from different perspectives can provide new research paths. From my own experience working alongside psychologists, then biologists, and now also mathematicians, engineers, and phoneticians, everyone approaches the same problems a little bit differently, and sometimes we surprise each other about which aspects of a problem we think of as difficult and easy.
Dr. Marisa Hoeschele, Priv. Doz.
Marisa Hoeschele conducts cross-species research on how humans compare to other species in their perception, production and appreciation of sound.
Academic Background: Marisa Hoeschele got a B.A. in Psychology at the University of Guelph, Canada in 2006. After that she completed an M.Sc. and PhD in Psychology with a specialization in Comparative Cognition and Behaviour at the University of Alberta, Canada in 2013. In 2013 she moved to Vienna as a post-doc and built the budgerigar laboratory at the Department of Cognitive Biology at the University of Vienna. In October 2018 she moved to the Acoustics Research Institute to build the “Musicality and Bioacoustics” research group where the budgerigar laboratory now resides. In December 2022 she received her Habilitation from the University of Vienna
