People often describe cats as a bit inscrutable and hard to read; in part, this is because cats have fewer facial muscles and fewer facial expressions. The expressions they do make are not necessarily like those of humans, so it may be challenging for humans to read them!
Put on your scientist hat for a moment: how would YOU measure cat facial expressions? And what would you imagine might be some challenges?
Why facial muscle movements matter
Some recent studies focused on quantifying the types of facial movements cats can make, and how they might be correlated with emotional or physical states, such as pain. Among many species, including mice and rabbits, facial expressions have been highly correlated with are an indicator of physical state, often through changes such as a wince, a tightening around the mouth or eyes, and changes in the position of whiskers. For example, in 2019, the Feline Grimace Scale was validated as a quick and easy way to categorize whether cats were experiencing pain.
How can we measure facial expressions in cats?
To measure facial movements in our cats, we have to do more than just gaze at them lovingly! Science to the rescue!
With a very long title, “The Application of Geometric Morphometrics to Explore Potential Impacts of Anthropocentric Selection on Animals’ Ability to Communicate via the Face: The Domestic Cat as a Case Study,” this research first looked at how cats’ facial features differed by face type. The three groups of cats were brachycephalic (with a broad, short skull, such as British Short-haired and Devon Rexes), dolichocephalic (with a long skull, such as Sphynx and Abyssinians), and mesocephalic (not too long or short, Goldilocks would say, just right). Although fewer cats are bred than dogs, breeders select cats for “extreme” features, which is how we ended up with short and long skulled cats in the first place.
Study 1: Comparing facial landmarks by breed and face shape
The purpose of the first study was to identify physical differences between cats based on breed and group (meso-, brachy- or dolicho-cephalic). To identify facial muscle movements in cats, scientists can use the Cat Facial Action Coding System (CatFACS). The researchers looked at 1888 photos from 19 different breeds of cats with “neutral expressions” (as agreed upon by multiple experts). By marking 48 specific facial landmarks based on the facial action units from the CatFACS, the researchers were able to determine three distinct categories of differences between the groups.
To summarize their results: there were differences between face types in the distance between the ears and different ear landmarks, how close the nose and eyes are to one another, and how spaced out the cheeks and mouth are. There were differences between the three groups, and also some differences between different breeds of cats. Mesocephalic cats were similar, and dolichophallic breeds showed the biggest range of variability in features.
Study 2: Real world implications
Differences between these groups are not too surprising, since cat breeding selects for physical characteristics. It’s what happened next where things get interesting.
The researchers first analyzed photographs of 25 domestic short-haired (DSH) cats after spay surgery, from both before and after the cats received pain control. Cat faces were analyzed as in Study 1, and each photo was also assigned a pain score. To compare, 25 photos from Study 1 (neutral cat faces of different breeds) were also assigned a pain score.
Results: Does your face identify pain if you are purebred?
For domestic shorthaired cats, there were differences in pain scores based on whether the cat had yet received pain medication. Among cat breeds, those with brachycephalic faces scored for more pain on the pain scale. Cats with “neutral” flat-faces have facial features and muscle positioning similar to that of a DSH in pain. Conversely, long-faced cats, Sphynx in particular, had fewer pain-like features.
This means that pain signals are less reliable in some purebred cats, especially if a pain scale developed on mesocephalic cats is used. It is unknown whether these effects would extend to other indicators of pain, such as body postures or vocalizations. But there are many questions – do cats whose faces naturally signal for pain increase caretaking behavior from their owners? Do Sphynx have a hard time letting their human know when they aren’t feeling well? We also don’t know if the facial indicators of pain are because those cats were actually in pain. Since the source of photos was from the internet, the actual state of the assessed cats is unknown.
Many purebred cats are prone to health problems due to genetic bottlenecks. Brachycephalic cats in particular suffer from dental malformities, difficulties breathing, and neurological problems. Changes in how they communicate with humans (and possibly other cats) may be an additional side effect of poor breeding practices.
It should be noted that the effects in this study were small, and the results were at times messy, meaning that there was a lot of overlap between groups, and we should interpret the results with caution. Ideally, future studies can explore whether the faces of purebred cats have musculature that masks or enhances signals for pain.
Finka, L. R., Luna, S. P., Brondani, J. T., Tzimiropoulos, Y., McDonagh, J., Farnworth, M. J., … & Mills, D. S. (2019). Geometric morphometrics for the study of facial expressions in non-human animals, using the domestic cat as an exemplar. Scientific reports, 9(1), 1-12.
Finka, L. R., Luna, S. P. L., Mills, D. S., & Farnworth, M. J. (2020). The application of geometric morphometrics to explore potential impacts of anthropocentric selection on animals’ ability to communicate via the face: the domestic cat as a case study. Frontiers in Veterinary Science, 7, 1070.
Schmidt, M. J., Kampschulte, M., Enderlein, S., Gorgas, D., Lang, J., Ludewig, E., … & Ondreka, N. (2017). The relationship between brachycephalic head features in modern persian cats and dysmorphologies of the skull and internal hydrocephalus. Journal of veterinary internal medicine, 31(5), 1487-1501.