Eavesdropping on speech in wild carrion crows

Large-billed crows wild-caught in Tokyo respond more to playback of Dutch than to Japanese without any prior training or rewards, but it remains unclear why they would respond differently to different languages. It is conceivable that they eavesdrop on human speech as an indicator to human presence (similar to eavesdropping on predator calls) and have subsequently habituated to Japanese, the locally spoken language. In this thesis, I aimed to investigate this response to speech in another corvid species, the carrion crow (Corvus corone). I found that wild carrion crows in both London and Milton Keynes responded to speech with increased vigilance compared to parakeet and pigeon calls (non-speech control vocalizations), but they did not respond differently to the local English compared to the foreign Vietnamese. To test whether this is due to a lack of ability or a lack of relevance, I then conducted a learning experiment in which playback of Vietnamese was paired with a food reward, while English was not. The crows in this experiment did not learn to discriminate the two languages. To understand whether encounters with humans are primarily positive, neutral, or negative for crows, I asked London and Milton Keynes residents about their behaviour towards crows. I found that while most people do not engage with crows, a small minority do regularly feed or chase them. I also surveyed carrion crows’ foraging locations across Milton Keynes to examine under which environmental conditions they may benefit from eavesdropping on speech. Finally, I wrote an agent-based model to understand whether eavesdropping could improve the crows’ survival odds, thus potentially providing a fitness benefit. The findings presented in this thesis contribute to research on corvid cognition, speech processing in non-human animals, and urban ecology.