Can animal studies be truly objective?
In the early 20th century, a horse in Berlin captured the world’s attention with its mathematical proficiency. When asked an arithmetic question, Hans, the horse, would start tapping his hooves and stop when he arrived at the correct number. This remarkable feat elevated Hans’ status to genius.
A psychologist named Oskar Pfungst, however, sought more than just a tacit acceptance of the horse’s intellectual prowess. Through carefully crafted experiments, Pfungst proved that Hans was responding to subtle, microscopic cues from the face of the questionnaire. The hoof taps were in response to involuntary signs from the human. The “Clever Hans effect” thus was the result of human bias in the interpretation of animal behaviour.
Eliminating bias
There is an important lesson here – an acknowledgement that we need to observe animals objectively and avoid projecting human emotions or motivations on them. In the last century, considerable efforts have been made to design experiments that can eliminate the inherent bias of the researcher.
“We try to base our conclusions on observed behaviours rather than subjectively assumed mental states,” says Rohini Balakrishnan, Professor at the Centre for Ecological Sciences (CES), who studies crickets and bats. She feels that even seemingly neutral terms can be misleading. “When we talk about an animal’s behaviour as a strategy, it’s easy to imagine the animal consciously evaluating its actions,” she explains. “But in reality, we’re just studying whether these actions lead to survival or reproductive success.”
In studying animal behaviour, anthropomorphism – attributing human qualities to animals – can be a significant challenge.
In studying animal behaviour, anthropomorphism – attributing human qualities to animals – can be a significant challenge.
“When humans and monkeys perform the same tasks, they often behave differently. These differences can easily be misinterpreted if we do not account for the species’ unique ways of perceiving or making decisions,” notes SP Arun, Professor at the Centre for Neuroscience (CNS) whose area of focus is visual perception in humans and monkeys.
On the field
Even in a carefully designed experiment, biases remain a persistent challenge, particularly in fieldwork where it is difficult to control the environmental and social variables. Rohini recalls a time when her team observed female crickets seemingly “rejecting” certain males during mate selection.
“We interpreted this behaviour as a rejection,” she says, “but when my former student ran simulations, we realised that it could simply be a result of how the females localise sound in a complex acoustic environment where multiple male calls overlap and mix.”
This discovery is a reminder of how our confirmation bias can colour the interpretation of animal behaviour, even when a simpler explanation may exist.
Saskya van Nouhuys, Associate Professor at CES, who studies wasps that parasitise other insects by laying eggs in them, recollects how bias can even shape the research trajectories. “We assumed that a wasp species, being large, parasitised butterfly caterpillars, but after years of failed experiments, I observed them parasitising the eggs instead,” she explains.
This observation was remarkable. The wasps were laying eggs inside the butterfly eggs. When these butterfly eggs hatched, the wasps developed inside the growing caterpillars and later emerged as adult wasps.
“The animal’s size is such a good reference that nobody would even question it,” Saskya adds; this breakthrough was a result of simple observation of the natural environment.
Many field researchers spend a lot of time observing animals in their natural habitats to combat bias before designing experiments. “Before conducting field experiments, my students often spend entire seasons just watching the animals,” says Rohini.
“That gives immense insights into which manipulations may work, and which might fail, increasing the chances of designing successful experiments.”
Bias also influences the interpretation of data and results. At this stage, Saskya emphasises improving communication between her lab members. “Do not be shy about discussing data with people; it can be students, technicians, or other colleagues,” she recommends. “And do not be too attached to your expectations.”
The roots of bias
Although cognitive biases may seem like a problem in studying animal behaviour, they are an inherent part of human existence, one with a deep evolutionary origin.
Sridharan Devarajan, Associate Professor at CNS, who studies attention and decision-making, believes that cognitive biases have an adaptive value. The constant deployment of attention to the external world requires significant cognitive effort. However, by relying on biases based on our past experience, our brain can optimise where and how much to pay attention to. For example, when driving on a familiar route, we may know that some junctions are less crowded than others; this allows us to bias our attention towards the busier junctions. This evolutionary perspective suggests that biases, such as expectation bias or pattern recognition, help animals navigate dangerous environments.
“That’s likely why many of the prey animals move as herds because each animal can monitor some part of the environment and raise an alert if something happens there. Predatory animals are the opposite; they do not need to put their attention all over the place. They focus on exactly one out of a few areas because that’s where the most vulnerable prey is, and they have to track and catch that one prey,” he informs.
Bias as a tool for interpretation
Not all researchers, however, see bias as a problem that needs to be eliminated. Anindya Sinha, a professor at the National Institute of Advanced Studies (NIAS), Bangalore studying primates in the wild, believes that our inherent biases are a crucial component in scientific thinking. The training of a researcher often defines the class of questions they prioritise. “If you are a geneticist, you could look at a problem in terms of gene function. If I were a biochemist, I could look at perhaps the chemistry of gene expression. And if I am an animal behaviour researcher, I would possibly look at the phenotype produced by these genes,” Anindya explains. “Our biases, on their own, are value-neutral. So, it has to be decided contextually whether a certain bias is helping us in our science or taking away and obscuring what is out there.”
Anindya advocates for reevaluating the approach towards studying animal behaviour. “There is a philosophical divide, represented by the natural and social sciences. When we say natural sciences, the laws of physics, mechanics, and mathematics come to mind. I think it has been a fundamental problem that we have used this philosophy of science, as encompassed by physics or mechanics, to colour our study of biology.”
When we deal with individuals and societies, we cannot directly apply the laws of physics, chemistry, and mathematics, according to Anindya. These are complex systems, with a variety of factors impacting single behaviour.
“Whether a monkey will eat some fruit depends on a range of factors. These factors could be biological, such as age and gender. It can be ecological – are there competitors around? Even more importantly, it could be experiential factors, maybe the monkey did not like the taste of the fruit yesterday. All these come together and determine whether a monkey will eat a fruit or not.”
In such situations, interpretations depend on training, akin to the story of four blind men describing an elephant and doing so differently. While none are wrong, the explanations stem from their individual differences. “All our views and assumptions are, in the end, our biases, revealing themselves,” says Anindya.
‘All our views and assumptions are, in the end, our biases, revealing themselves’
One’s biases are an integral part of learning and understanding the world. Although one must seek to minimise their prejudice to gain an understanding of natural phenomena, there is a need to find a balance.
“You take away all the biases, all the interpretations go away as well,” adds Anindya.
To study animals with complete objectivity has been an ideal pursued by many. However, biases do have a role in enriching scientific interpretations. “It is impossible to think of science, whether of animal behaviour or of anything, without factoring the biases which have helped us. As long as we recognise, expose, and clarify our biases, they can promote healthy scientific discourse,” Anindya stresses.
Vaibhav Sharma is an Integrated PhD student in the Department of Microbiology and Cell Biology, IISc and a science writing intern at the Office of Communications
