Reaction to “Embracing Anthropomorphism in Canine Care: A Paradigm Shift for Professionals”

This is my reaction to the recent article titled “Embracing Anthropomorphism in Canine Care: A Paradigm Shift for Professionals” from an entity called NICE.  Let’s start with the opening claims that anthropomorphism has recently been “increasingly recognised as a valuable tool in enhancing the care, training, and understanding of dogs” ; contrasted with it previously being  “considered an unscientific approach”.  Both parts of this two-part claim are misleading.  First, we know that freely and uncritically interpreting the behaviors of animals anthropomorphically is in fact unscientific and often detrimental to animal welfare.  Examples abound.  Here are just a few:

  1. The entry on anthropomorphism in the Oxford Companion to Animal Behavior (McFarland 1987) recounts the widespread, but erroneous belief at the time, that the first chimpanzee into space enjoyed his flight because he returned to earth “grinning broadly”. In reality this facial grimace is an expression of extreme fear.
  2. An example we are all familiar with is pet owners describing their dogs’ “guilty looks” when they arrive home to evidence of misbehavior. Yet we know the behaviors they observed are a combination of the expression of fear and submission.
  3. In lectures years ago I used the example of research done with sheep – I can no longer find the article or remember all the details. But the authors had subjected sheep to several stressful conditions – being separated from their flock, being sheared, and I believe being transported in a truck.  When asked, most people thought being sheared would be the most stressful, yet the highest cortisol levels occurred during social separation from the flock.
  4. I can recount a personal, anecdotal example from our days owning a self-service dog wash. One owner said she knew her dog enjoyed his bath and blow dry because he always yawned, which showed how relaxed he was.  The yawning as we all recognize is a displacement behavior, not a sign of relaxation.

We can all think of hundreds more examples, none of which would support the contention that anthropomorphism is a “valuable tool” enhancing animal welfare.

Second, the article implies that scientists have only recently taken a second look at anthropomorphism.  This is untrue and shows an unfamiliarity with the last 40 years of behavioral literature.   It’s true that behaviorists and ethologists have long cautioned against the dangers of anthropomorphism, because it easily leads to misinterpretations of behavior and can be detrimental to animal welfare.  Such concern is, and continues to be, justified – just consider the above examples.

Third, scientists have also talked about the use of “critical anthropomorphism” – a term introduced by Gordon Burghardt in the mid 1980s (see  One way to think of “uncritical anthropomorphism” is, “if I was in situation X this is how I would feel, and because my dog is just like me, she would feel/behave the same way”.  Compared with “critical anthropomorphism” which could be described as something like “if I were a social predator with exquisitely developed senses of smell and hearing, predisposed to cooperate with conspecifics and avoid conflict, I would be likely to show behaviors ABC in situation X”.  In other words, one takes into account the biology, physiology and natural history of the particular animal.

Burghardt goes onto say that “critical anthropomorphism” can help scientists ask better questions and design better research.  In fact, we as humans cannot completely avoid being anthropomorphic  in our interpretations about animals and their behavior because we can never ever know directly exactly what the experiences, or the lives of other animals are like.  We can’t even know that about other humans!

Lehner (1996) in the second edition of Ethological Methods discussed anthropomorphism as being along a gradient rather than its presence or absence.  At one end of the gradient is no anthropomorphism (referring to how behavior is described such as ears back, tail down); followed by the use of human terms technically defined (an example from authors Estep and Bruce regarding defining “rape” ethologically as forced copulation or resisted mating to prevent the association with connotations  from the human context); next using human terms as a metaphor (behaving “as if” followed by a human attribute or emotion) and lastly human terms used freely without restriction or qualification – as in the guilty looks example.

The article argues that because neuroscience has now found similarities between human and canine brain functions, that we now know that they experience emotions that humans experience like “joy, fear, and anxiety”.  But do we really know that?  We do not. The neuroanatomy provides evidence for the capability for existence of those emotions (the denial of which has long since passed) but do we really know they are the same as ours?  Perhaps dogs or other animals feel emotions more intensely, or less intensely than we do, and in different contexts.  How can their experiences not be different from ours when their sensory systems and ours have such vastly different capabilities?  They can hear things we can’t, see things we can’t, smell things we can’t, or at least better than we can, and vice versa.   In fact other recent research (I don’t have the citation right at hand) has found connections in the brain between the visual areas and olfactory areas in dogs that don’t exist (or at least to the same degree) in humans.  And if we stop and think about it, examples again abound of how animals respond differently than us in many situations in part at least because they perceive them differently (easy and obvious – their penchant to roll in smelly things).

But back to recent findings from neuroscience.  Acknowledging the similarities that have been found between humans and dogs in certain brain mechanisms and anatomy is not anthropomorphism but instead a reflection of research findings.  It’s not anthropomorphic to state similarities exist.  But we should NOT forget that differences also abound, and we ignore them at the animals’ peril.

The article goes further to describe anthropomorphism as a “methodology” or “tool”.  One definition of method is “a system of principles for doing something”.  Anthropomorphism is NOT a “system of principles”.  It is a perspective of interpretation.  I for one do not want my approach/ the methods I choose to use for behavior modification and training to be based on anthropomorphism.  I want my chosen methods to be based on science and what is best for the animal.   I don’t think the field of dog training and applied behavior has come to focus on the use positive reinforcement because it is more “empathetic”.  I think it’s because its use (and not always its exclusive use) is what’s most effective and best for the animal.

Allowing new scientific findings about brain function and emotions to inform our decisions about how to improve our approach to training and behavior modification and therefore improve animal welfare I would argue is not being anthropomorphic but rather a natural progression that applies the findings of research to real world issues – something that has always been part of scientific research.

In summary, I think while perhaps the article is well meaning, it is flawed.  It reveals a lack of familiarity with the breadth of discussions about anthropomorphism, including its dangers and potential usefulness, that have been part of the scientific literature for decades; it confuses a straightforward acknowledgement of scientific findings that demonstrate certain similarities between humans and dogs with anthropomorphism; it erroneously refers to anthropomorphism as a methodology; it claims that anthropomorphism is “essential” to create compassion which ignores the possibility that one can have deep appreciation and compassion for animals stemming from their very uniqueness and differences from people (how many times have we heard dogs give unconditional love, dogs don’t judge us, etc.) and it ignores the “downsides” or potential dangers of anthropomorphism that we’ve all seen manifest time and again.

Final thoughts on the matter I had during my run this morning:  In conclusion, maybe the best we can do is to expand on Lehner’s model and consider the interpretation of animal behavior along a continuum  – from  as purely close to scientific, absent anthropomorphism as possible, to relying on interpretations based solely on a human-centric view of animals.  Recognize that scientific interpretations do not preclude empathy and compassion. The days of Descartes and seeing animals as robots devoid of emotions are long gone.  It is the rare behavior scientist (I don’t personally know any) who deny animals have emotions, but they do continue to legitimately question the nature of those emotions – and that’s a good thing, it’s what drives research.  If we rely mostly on anthropomorphism, then we are back to the days of guilty looks, “he knows he’s done wrong”, and chimpanzees who enjoy space flight, to the detriment of animal welfare.  It is my opinion that the article “Embracing Anthropomorphism in Canine Care: A Paradigm Shift for Professionals” misses the point, and further muddies the waters of interpretation of behavior rather than serving to clarify them and make a positive contribute to our understanding.

The article is attractive because it appeals to the strong feelings we all have about dogs and our passion to do what is best for them.   But as always if we are continually striving to be “science based” then we must bring that, and our critical thinking skills to the analysis of this article.