Can animals volunteer to participate in research? If so, what does volunteering look like, and what does it mean for animal welfare?

'We agreed that to be truly voluntary, participation in research should ideally be inherently rewarding in itself...'

Puzzling over these complex questions has been a popular pastime within the Animal Research Nexus team. Beth and Emma have explored what ‘informed consent’ means, and could mean, in research with both human and animal participants (Greenhough and Roe 2011). Vanessa and Pru have considered similar issues around consent in veterinary medicine, such as in the case of canine blood donation (Ashall 2017; Ashall et al. 2018; Gray et al. 2018).

For my part, during my PhD research on orangutan conservation I was confronted by a disagreement about the meaning of ‘choice’ for animals (Palmer 2018). When attempting to send an orphaned orangutan who has been kept as a pet back to the forest (a process known as rehabilitation), one might find that the orangutan doesn’t particularly want to give up the comfort of regular feeding and human contact. So rather than become ‘wild’, the orangutan might keep coming back to the release camp to take advantage of food supplied by the release crew, or (sometimes aggressively) approach any visitors allowed at the site in the hope they’ll offer food. This can even become an inter-generational strategy, with orangutan mothers teaching their children to forego regular foraging for a human-supplied meal.

While one rehabilitation practitioner I spoke with emphasised that it’s the orangutan’s choice whether they want to become ‘wild’ or continue living alongside humans, others described ongoing human dependence as a product of training. Rijksen and Meijaard (1999, p.161) have gone so far as to say that rehabilitated orangutans who remain dependent on human care, protection, and food are in ‘a peculiarly enslaved situation’ (rather the opposite of a free ‘choice’!).

The ‘lower threshold’

This tricky question of what it means for an animal to ‘choose’ has come up once again in my current research for the AnNex project, which focuses on what happens when animal research moves out of the lab and into the wild, farm, fishery, zoo, or vet clinic. Technically speaking, I’ve been looking at activities that are regulated by the Animals (Scientific Procedures) Act (A(SP)A) of 1986, which was initially written with lab animals in mind (an early draft was called the ‘Laboratory Animals Protection Bill’), but happen in ‘places other than licensed establishments’ (POLEs).

POLEs projects tend to not be particularly invasive – a typical project might involve catching, blood sampling, and releasing a wild animal to collect data for answering various ecological questions. POLEs projects are therefore often rather borderline when it comes to whether or not they fall under A(SP)A. The important criteria for inclusion under A(SP)A are that the activity: 1) uses ‘protected’ animals (Reuben is looking at arguments around inclusion in this category); 2) is done for research rather than veterinary treatment or ordinary husbandry; and 3) causes ‘pain, suffering, distress or lasting harm equivalent to, or higher than, that caused by inserting a hypodermic needle according to good veterinary practice’, known as the ‘lower threshold’ (ASRU 2016).

The level of suffering caused by various research activities I have studied has been subject to debate. For example, how heavy can a tracking device be before it starts to seriously affect a bird’s welfare (Wilson 2017)? Another consideration is whether the animal is accustomed to the procedure. For example, what if an animal housed in a zoo becomes so accustomed to providing blood samples that the activity no longer causes them any significant stress?

Using this reasoning, Goodrowe (2003, p.320) has suggested that in the context of US zoos, once a behaviour becomes ‘reward based and voluntary’, it may be considered ‘standard animal management practice (vs. research protocol)’. In other words, if an animal ‘volunteers’ for something like a blood sample, it should no longer be considered invasive research but just ordinary husbandry.

‘Voluntariness’ can to some extent affect whether an A(SP)A licence is required. For example, radio collaring or microchipping an animal may not require a licence if this can be done without anaesthesia (an easier task with a zoo animal who ‘volunteers’ than a wild animal), but becomes licensed if anaesthesia is required (ASRU 2016).

But training and ‘voluntariness’ only work up to a point, due to an issue with loopholes: if ‘voluntary’ blood samples aren’t regulated by A(SP)A and non-voluntary samples are (on the grounds that the former is not very stressful), there might be an incentive for researchers to misread the situation. This is because – as I’ve heard from many research participants – securing an A(SP)A licence can be time-consuming. I’ve also been told that institutions like zoos are often concerned that getting an A(SP)A licence, which is intended to regulate invasive research with animals, would have negative public relations consequences. An animal’s level of stress also depends on contextual factors like the behaviour and skill of the person taking the blood sample, so it could be difficult to say consistently how stressful an animal finds a procedure.

The hypodermic needle line, or ‘lower threshold’, therefore acts as a kind of blanket rule within A(SP)A. Plenty of researchers I’ve spoken with have taken issue with this rule, since the use of a needle means very different things for different animals, and there are potentially practices that are more stressful for the animal than a needle that aren’t covered by A(SP)A (various methods of trapping wild animals being commonly cited examples). However, the loophole problem indicates that there might also be a pragmatic justification for this policy. As with the five-day rule, which determines when zebrafish count as animals under A(SP)A, the hypodermic needle line could potentially also be described as ‘arbitrary but not indefensible’.

Defining and assessing voluntariness

Even if it isn’t practical to consider voluntariness in assessing whether research falls under A(SP)A, voluntariness could still have important implications for animal welfare. But how do we know whether an animal ‘volunteers’ for research?

To help answer this question, I got in touch with Professor Hannah Buchanan-Smith of the University of Stirling, whose research, among other things, involves both using and assessing voluntary methods for non-invasive cognition research with zoo-housed primates. Together we decided to discuss questions around voluntary research with a group of staff and students from the Behaviour and Evolution Research Group (BERG) at Stirling in January 2019. The group worked with a diverse range of species and contexts, from chimpanzees and elephants in zoos and the wild, to crows in the ‘wild’ of the university campus, to dogs in animal-assisted therapy.

Professor Buchanan-Smith and I had joked before the meeting that the most frustrating conclusion is always, ‘it must be assessed on a case-by-case basis’. Our jokes must have jinxed the group discussion, as that was essentially our conclusion.

In terms of a definition of voluntary research, we agreed that it’s important that there’s no coercion. This means that an animal’s alternatives to participating in the research must be pretty desirable. For example, a member of the group told us about a case of doing non-invasive cognition research with a horse in a lush field of grass – the field was surely not a bad alternative if the horse wasn’t interested in the research. Contextual factors are therefore crucial, with animals in more restricted environments (Gazes et al. 2013), and with certain personalities and life histories (Herrelko et al. 2012; Morton et al. 2013), being more willing to engage in research.

But our definition was complicated by a few outstanding questions posed by the group. For example, if an animal is already in a captive and controlled environment, can their participation ever be truly voluntary? In animal-assisted therapy for example, the dogs indicate if they want to leave, but that’s after they’ve already been taken to the therapy location in the first place.

We were also puzzled by the boundaries between volunteering and bribery, the key questions being, how much does the animal not want to do something, and how much is being offered to convince them? The quantity and quality of food rewards might therefore be important. However, as research by Frans de Waal (2005) has shown, quantity isn’t everything, since capuchins who were previously very happy to receive cucumber became upset and uninterested in cucumber when their peers received grapes, demonstrating the contextual nature of decisions (in this case inequality).

We agreed that to be truly voluntary, participation in research should ideally be inherently rewarding in itself. We might not just want to do this for ethical reasons, but also because an animal’s level of interest in research can alter results. Despret (2004) has explored how results can differ depending on whether animals have been given opportunities to be ‘articulate’ and show the researchers what is interesting to them, compared with research intended to make animals docile and ‘inarticulate’, which tells us more about the experimental design than the animal (see also Greenhough and Roe 2011). As Despret points out, animals don’t need to be interested in the research for the same reasons as the researchers; the point is that human-animal relationships and results change when animals find the research interesting for their own reasons.

Similar considerations about inherent reward and payment are at play with human research participants, hence why the NHS Health Research Authority (2014) offers guidance on the circumstances when payment for medical research might result in ‘coercion’ and ‘undue inducement’. In short, with both people and animals we worry that lucrative rewards might convince people to do things they wouldn’t otherwise do. (Then again, we might not: Goodrowe (2003) seems quite happy with the idea that animal participation in research can be simultaneously reward-based and voluntary.)

But defining volunteering as involving inherent reward rather than payment could be taken to mean that a large amount of what both humans and animals do isn’t truly voluntary, including work. Much has been written about the meaning of work for both humans and animals. For example, Clark (2014) has explored whether it makes sense to think about animals involved in clinical trials not as research tools but as labourers. According to Clark, the idea of animals as fellow labourers makes some sense once we acknowledge that animals are agents with purpose.

However, Clark also points out that ‘working’ animals might be, to varying degrees, coerced – they might be ‘significantly unfree partners’, to use Haraway's term (2008). This means that we need to look at ‘the human-animal relations of production’ on a case-by-case basis when thinking about an animal labourer’s level of autonomy. Haraway helpfully suggests that we might think about this in terms of variable ‘degrees of freedom’ experienced by research animals. In a similar vein, we as a group thought about the idea of voluntariness as a scale rather than a binary definition.

In terms of assessing voluntariness, consent forms are typically used for this purpose with human research participants (Greenhough and Roe 2011; Gray et al. 2018). Establishing consent is harder with humans who are judged not capable of giving informed consent, hence the use of the term ‘assent’ rather than ‘consent’ in research with children (Ashall 2017). Similar issues arise when seeking ‘assent’ from (non-human) animals.

One way of addressing this issue is to look closely at animals' behaviour and body language to get a sense of their preferences (Beth and Emma wrote about this in terms of ‘somatic sensibilities’). For example, the group discussed various behavioural cues used for assessing the preferences of primates at the Living Links project, such as when individuals simply walk away from the research activity, or put a hand on the door if they’re looking to leave a research cubicle.

We weren’t sure that it’s necessary for animals to understand the purpose of the research, since we humans might volunteer for things that involve an element of surprise. But perhaps we want to tell animals about the risks – as Vanessa and Pru have pointed out, the principle of informed consent is based on the idea that humans should have choice over the degree of risk they’re willing to put themselves through for research or medical treatment.

But accurately communicating risks to other animals is difficult, so perhaps we should instead make decisions on behalf of animals that we believe will best serve their interests? Vanessa and Pru have argued that this is an important principle when thinking about veterinary treatment, which would represent a break from the paradigm of allowing owners to make decisions on their pet’s behalf, irrespective of whether the decision is in the animal’s best interests (Ashall 2017; Ashall et al. 2018; Gray et al. 2018).

Other considerations

We agreed that it’s important to carefully consider how even voluntary research might negatively affect animal welfare. For example, removing the research activity could be stressful, as might returning an animal to its group after research (Ruby and Buchanan‐Smith 2015). We might therefore want to ask whether animals are better off not participating in the research at all, versus participating and having the research end. Similar issues can arise with research with humans, such as if beneficial treatments used in clinical trials are not available to research participants after the study ends (e.g., see Zong 2008).

We also discussed how voluntariness and animal welfare might not be the only values worth considering. Coming back to the orangutan rehabilitation example, we talked about how you might see encouraging orangutans to become ‘wild’ – in the sense being independent of humans – as valuable in itself. You might therefore conclude that, irrespective of whether taking food from humans is a free ‘choice’, it should still be prevented. Similar considerations could apply to research with wild animals – for example, we might want to think about the effects of using food rewards if this could habituate a wild population to human presence.

It’s probably unsurprising that our discussion group wasn’t able to answer all these questions with any certainty. Ideas about voluntariness, choice, and work affect not only how we think about animal research, but also how we think about our own actions. After all, as a participant in my PhD research observed, if I see orangutans as ‘trained’ to rely on human provisioning, then I have to acknowledge that I was ‘trained’ to be a PhD student! I prefer to think doing a PhD was voluntary, but then again I haven’t been able to tell you what ‘voluntariness’ actually means.


I’d like to sincerely thank everyone who attended the discussion group from BERG at the University of Stirling for your help in thinking through these issues. Particular thanks go to Professor Hannah Buchanan-Smith for organising the meeting, and for recommending references on this subject.



Animals in Science Regulation Unit (ASRU). 2016. Animals (Scientific Procedures) Act 1986: Working with Animals Taken from the Wild. Home Office.

Ashall, Vanessa. 2017. Veterinary Donation: To What Extent Can the Ethical Justifications for Living Human Donation Be Applied to Living Animal Donation? PhD thesis, Nottingham: University of Nottingham.

Ashall, Vanessa, Kate M. Millar, and Pru Hobson-West. 2018. ‘Informed Consent in Veterinary Medicine: Ethical Implications for the Profession and the Animal “Patient”’. Food Ethics 1 (3): 247–58.

Clark, Jonathan. 2014. ‘Labourers or Lab Tools? Rethinking the Role of Lab Animals in Clinical Trials’. In The Rise of Critical Animal Studies: From the Margins to the Centre, edited by Nik Taylor and Richard Twine, 139–64.

De Waal, Frans B. M. 2005. ‘How Animals Do Business’. Scientific American.

Despret, Vinciane. 2004. ‘The Body We Care for: Figures of Anthropo-Zoo-Genesis’. Body & Society 10 (2–3): 111–34.

Gazes, Regina Paxton, Emily Kathryn Brown, Benjamin M. Basile, and Robert R. Hampton. 2013. ‘Automated Cognitive Testing of Monkeys in Social Groups Yields Results Comparable to Individual Laboratory-Based Testing’. Animal Cognition 16 (3): 445–58.

Goodrowe, Karen L. 2003. ‘Programs for Invasive Research in North American Zoos and Aquariums’. ILAR Journal 44 (4): 317–23.

Gray, Carol, Marie Fox, and Pru Hobson-West. 2018. ‘Reconciling Autonomy and Beneficence in Treatment Decision-Making for Companion Animal Patients’. Liverpool Law Review 39 (1): 47–69.

Greenhough, Beth, and Emma Roe. 2011. ‘Ethics, Space, and Somatic Sensibilities: Comparing Relationships between Scientific Researchers and Their Human and Animal Experimental Subjects’. Environment and Planning D: Society and Space 29 (1): 47–66.

Haraway, Donna J. 2008. When Species Meet. Minneapolis: University of Minnesota Press.

Herrelko, Elizabeth S., Sarah-Jane Vick, and Hannah M. Buchanan-Smith. 2012. ‘Cognitive Research in Zoo-Housed Chimpanzees: Influence of Personality and Impact on Welfare’. American Journal of Primatology 74 (9): 828–840.

Morton, F. Blake, Phyllis C. Lee, and Hannah M. Buchanan-Smith. 2013. ‘Taking Personality Selection Bias Seriously in Animal Cognition Research: A Case Study in Capuchin Monkeys (Sapajusapella)’. Animal Cognition 16 (4): 677–84.

NHS Health Research Authority. 2014. HRA Ethics Guidance: Payments and Incentives in Research. April. Vol. V1.0.

Palmer, Alexandra. 2018. Saving and Sacrificing: Ethical Questions in Orangutan Rehabilitation. PhD thesis in anthropology, London: University College London.

Rijksen, H. D., and E. Meijaard. 1999. Our Vanishing Relative. Dordrecht: Kluwer Academic Publishers.

Ruby, Suzanne, and Hannah M. Buchanan‐Smith. 2015. ‘The Effects of Individual Cubicle Research on the Social Interactions and Individual Behavior of Brown Capuchin Monkeys (Sapajus Apella)’. American Journal of Primatology 77 (10): 1097–1108.

Wilson, Rory. 2017. ‘Tags on Birds - How Much Are Our Guidelines Flights of Fancy?’ Presented at Norecopa: Harmonisation of the Care and Use of Wild and Domestic Mammals and Birds in Field Research, Oslo, 26-27 October.

Zong, Zhiyong. 2008. ‘Should Post-Trial Provision of Beneficial Experimental Interventions Be Mandatory in Developing Countries?’ Journal of Medical Ethics 34 (3): 188–92.


Nestlé, 'Cognitive testing' (licence here).

William H. Calvin, 'Bonobos Panbanisha & Kanzi with Sue Savage-Rumbaugh, 2006' (licence here).