“Are Those Cows Jumping For Joy?” Eye Whites, Hoof Kicks And Qualitative Behaviour Assessment

“Are Those Cows Jumping For Joy?” Eye Whites, Hoof Kicks And Qualitative Behaviour Assessment

Maisie Tomlinson

Even the social lives of urban citizens are always intimately entangled with other animals, silently circulated in prescriptions and sandwiches and lattes. Behind such everyday artefacts are the lives of multitudes of individual creatures, often in factory farms and laboratory facilities, whose lived experience of these systems seems to be of concern to increasing numbers of people. Public pressure over animal welfare helps shape the nature of production, if often in frustratingly small increments. But decisions about why and how to improve the lives of these animals can often be traced back to the production of scientific knowledge about them – what their physical and emotional needs are, and how viscerally they may be affected by the systems in which they are confined. Intrigued by how such knowledge is produced, and in what circumstances, my research into animal behaviour expertise has demonstrated that animal science comprises various and sometimes competing schools of thought. One interesting source of tension is that between an “objectivist” methodology—isolating  details of an animal’s behaviour whose relevance is grounded in evidence from controlled experiments involving quantitative comparisons—and that which comes from a more qualitative, intuitive, ‘tacit knowledge’. Whilst both may be used in welfare assessment, typically only the first is considered legitimate expertise, on which animal welfare’s hard-won acceptance as a science was eventually built. The second has a more ambivalent status, acknowledged for its role in effective animal management, but also risking the charge of “anthropomorphism’, a word usually used to claim the misattribution of qualities to animals that are presumed to be exclusively human.

One example which demonstrates the significance of this tension is the question of whether Cows Belong in Fields, to borrow from a prominent NGO campaign. The question of how cow welfare may be affected by the rise of intensive ‘zero-grazing’ systems for dairy cattle, where hundreds or even thousands of cows are housed permanently indoors in order to increase milk production and profitability, can influence decisions which  shape cows’ lives, the countryside, and human livelihoods for decades to come. Both proponents and opposers of such ‘mega-dairies’ use objectivist scientific evidence of cow welfare in their arguments. But both sides also rely on qualitative, intuitive interpretations when presenting cow preferences as self-evident. Some years ago I worked on a campaign to stop the spread of mega-dairies across Europe. In the Netherlands, at the beginning of spring there is a traditional event called the koeiendans’ (cow dance). Crowds of families gather to cheer on the release of the cows from their winter barns onto spring pasture. Typically they gallop onto the grass where they gambol, buck, and roll on the ground, giving every impression of delight. For an NGO, the ‘cow dance’ is a powerful campaign tool—a feel-good visual that poignantly reminds people of what is lost when you confine cows indoors—animal joy. Grazing, they say, isn’t just nutrition – it’s stimulation and exercise. It’s variation in smells and tastes and textures and weather conditions. It’s a choice over who you butt heads with. Watching the release of his own cows, farmer Neil Darwent, a prominent UK advocate of pasture systems, says:

“We do say cows belong in fields, and you can see quite clearly today that there are a very happy bunch of cows here”

However, an in-house scientist, watching a similar video, recognised the enlarged eye-whites of some of the cows as an established, measurable indicator of stress. This was evidence, he said, that the release onto pasture might be a negative experience. The running and bucking could be in fear. The organisation prided itself on evidence-based campaigning. To ‘anthropomorphise’ was explicitly defined as attributing emotions without scientific evidence. Despite a leading cow welfare scientist challenging the conclusion of the first, the videos were dropped from the campaign.

What struck me as strange was not that the cows could be stressed, especially by the noisy crowds and the sudden change of environment. Nor was it the possibility that their wide eyes might challenge our preferred interpretation of events. It was the assumption that an animal couldn’t be stressed and excited and happy and afraid, all at once. After all, aren’t we given to mixed feelings at new opportunities, at the firing of a start gun, at the Boxing Day sales? It’s not a given that other animals should be different. To assume that only humans have dynamic pallet of feelings whilst other animals are painted only in primary colours seemed reductive. And whilst we certainly can misunderstand the behaviour of other species, to assert that all judgements should flow only from those aspects we can measure, with everything else an untrustworthy potential ‘anthropomorphism’, seemed to deny millennia of successful interactions  as humans live and work with and hunt other animals.

This experience helped lead to my interest in a scientific animal welfare assessment methodology called Qualitative Behaviour Assessment. Initially developed for use in farm animals, but now extended to a range of contexts, QBA asserts that, as live subjects continually in response to their environment, animals’ emotional experiences are by nature dynamic and shifting, full of subtle details in expression. As such, it is argued, the most appropriate way of capturing and integrating this information is through the use of qualitative descriptions like ‘playful’, ‘sociable’ or ‘depressed’, terms which would often be considered ‘anthropomorphic’ in a scientific assessment. Its developer, Professor of Animal Science Francoise Wemelsfelder, is trained in philosophy as well as biology, and has mounted a substantial epistemological defence of QBA. She argues that uncertainty is part of any method, that the risk of errors can be mitigated by species-familiarity and sustained observation; and that quantitative methods too rely on ‘anthropomorphic’ theories taken from human psychology or economics. Whilst QBA is not designed to replace other assessments, she believes that incorporating emotionally resonant language has ethical effects, turning the animal from a collection of separate parts into a live being with meaningful emotional experiences. So that, she might say, when we watch a cow being released onto pasture, we need to take the “whole animal” into account simultaneously: the eye whites, the jumps, the tension in its body and the situational context – and integrate these using qualitative descriptive language. She has co-developed innovative statistical methods for comparing observer agreement and correlating QBA with physiological measures, and whilst some studies disagree, an independent scientific review of the QBA literature concluded that it was an effective methodology in combination with other techniques.

In my research, I wanted to explore some of the difficult questions that the use of QBA raised about the possibilities and limits of knowledge of other species, and the social and political context in which it arises. To this end I’ve been following a project in a UK laboratory animal facility which was developing and validating a species-specific list of QBA terms for the welfare assessment of mice. Talking to animal welfare scientists, vets and animal technicians, I discovered that, like claims about a cow’s need for pasture, claims made about the welfare of mice could have wide-ranging social implications, not only for the mice themselves, but for scientific careers, laboratory designs, the development of pharmaceuticals, and public opinion.

But it also highlighted local tensions about the different kinds of knowledge that could be legitimately marshalled in assessment. Some were sceptical about assessing mice using direct qualitative descriptions of their emotional state, like “frustrated”. Long experienced in misdiagnoses, including their own, they gave persuasive objections to confidence in ordinary observation. However, others expressed a growing interest in qualitative assessment, in the wider context of knowledge management and the cultivation of empathy. They argued that QBA merely formalised the kinds of ‘tacit knowledge’ that animal welfare professionals used all the time, even when ostensibly using physiological welfare indices like coat condition. They worried that over-reliance on pre-defined indices was risky, and thought that there was a legitimate role for intuitive judgement. Finally, some believed that even if the chosen qualitative descriptions were wrong, taking the extra time to choose the right words meant slowing down, integrating information, and appreciating the live-ness of the mouse:

“It might be just like a bridge, to looking at the mice in a slightly different way. Or just to be slightly more thorough… That actually, even if those aren’t really the right things to mark, the fact that you’ve taken longer to get there, you might get an overall impression, and you look at down and you say well, it says everything’s good but I’m not sure why I’m not happy with the mouse!” (Julian, laboratory vet)

In summary, animal welfare science can be instrumental in shaping the lives of millions of nonhuman others, but also the nature of work, of social institutions and of the countryside. As a relatively young science, vulnerable to censure, adamant ‘objectivism’ is understandable. But whilst precise physiological measurements remain indispensable, we should be cautious in being so fearful of the accusation of anthropomorphism that we revert to scientism, viewing such methods as the only way to understand the lives of nonhuman others. Otherwise we risk a reductive view of animals’ emotional experience which ultimately distorts our understanding. My research suggests that the growing acceptance of Qualitative Behavioural Assessment in a range of scientific settings is one way in which ‘tacit’, interpretive knowledge of animals is being reclaimed and made respectable. Along the way, fascinating questions are raised about the nature of knowledge and our relationship to other species, with real-life implications for both humans and other animals.


Maisie Tomlinson is a PhD student in Sociology at the University of Manchester. Her research explores the social politics of animal behaviour expertise as it is practiced in diverse settings. Shadowing people who study animals, her current work explores methods which assert a belief in the ‘innate’ ability of humans to read and interpret the behaviour of other species on some level. One is the teaching of horse behaviour in a rural organisation offering ‘Equine-Assisted Personal Development’; and the other explores a science-based animal welfare methodology called Qualitative Behavioural Assessment (QBA), following its development for use as a welfare assessment tool for laboratory mice at a UK university. Her work covers themes such as animal subjectivity, ‘anthropomorphism’, the role of science and the use of the body.

IMAGE CREDIT: © 2xSamsara.com / Adobe Stock