Distinguishing between fear and anxiety in dogs
Is it important to distinguish between fear and anxiety in dogs? Can we even distinguish between them? Does it matter?
Thinking about canine emotions can be very tricky.
On the one hand, if we think of our dogs experiencing the full range of emotions in exactly the same way humans do, we misunderstand the canine experience. It leads us to think of dogs experiencing guilt, shame and pride in very similar ways to humans. It also stops us thinking of them as dogs and understanding their own experience.
On the other, if we think dogs don’t have emotions, we’re likely to put them into situations without paying due care and attention to potential suffering. If we believe their only experience is physical pain, that can end up with us failing to pay attention to their emotional or mental distress.
The challenges of thinking about canine emotions
It can be hard to think in more nuanced ways about canine emotions.
Not only that, it tends to be a divisive subject in the dog world. Some trainers take certain positions and then adamantly defend them when it comes to a dog’s feelings. Even when defending one view or another, they can be very dogmatic in their understanding. This is normal, because emotion researchers tend not to explain that their field is incredibly divided and that there is little common ground.
If you only hear one or two people speak, and they do so authoritatively, you can be forgiven for thinking that there is only one way to see things.
This can then filter down to trainers. For instance, every time I see people delivering courses on emotional regulation, it’s a red flag to me that they don’t really have a firm grip yet on what they’re talking about. But such courses can be very popular and very influential in the dog world.
Why it is important?
Because the researchers exploring this topic call it emotion regulation. This is an important distinction. One suggests regulating with emotions and the other suggests regulation of emotions. Also, the more you read on the topic, the more you tend to assimilate the vocabulary they use, almost without thinking.
It’s really important we’re informed by knowledgeable and accurate individuals. They should really know their stuff.
Why it’s important to get the concepts right
It’s not just about misuse of key terms. When we misunderstand key concepts, we can also tend to present them in black and white.
Such courses also miss the subtleties such as the fact that emotion researchers are not entirely in agreement as to whether emotion regulation is even a thing or not.
Think of it as the dark matter of emotion research.
As Eysenck and Keane point out in Cognitive Psychology: A Student’s Handbook, there is no clear-cut distinction between emotion generation and emotion regulation.
The work of James Gross on emotion regulation is fascinating. It is the basis for a lot of Lighten Up materials on resilience for example. At the same time, there’s more than cognitive regulatory processes. We make behavioural choices that affect our emotions. We also make social choices that help regulate them too.
How good we are at regulating our emotions depends on the individual and our individual skills. It also depends on the situation. You might not be very effective at regulating your emotions in particular circumstances. It also depends on the effectiveness of the strategies we choose.
These are nuances that can be missed by people who are less familiar with the topic. When we accept concepts as information, as ‘facts’, then it can lead us to take strong stances on a topic without considering the nuance.
Dummond and Fischhoff (2017) explored the phenomenon that more knowledge can actually lead us to take more polarised positions. Ironically, the more we know, the more dogmatic and rigid our thinking can become.
Emotions as hardwired distinct processes
Many dog trainers and canine professionals refer to Jaak Panksepp’s book when thinking about canine emotions. In his books, he explored seven ancient mammalian emotional neural networks which he called affective pathways. He speculated that there were seven pathways shared by mammals: SEEKING, RAGE, FEAR, LUST, CARE, PANIC/GRIEF and PLAY.
What he researched was the neural ‘hardware’ underpinning emotions in subcortical structures in mammals. Although we now know that some of these pathways are much more complicated than he defined twenty years ago, and that they involve much more diffuse processes, Panksepp is often among the dog trainers’ favourites when it comes to explaining emotions.
Many people like the notion of emotions being distinct, clearly identified, involuntary, hard-wired processes with strongly physiological roots. It also put paid to some of the arguments over whether animals felt emotions at all. They certainly had the same hardwiring as humans do.
The challenges come in that some of these are not distinct processes in themselves, and many of the structures he studied are involved in a lot of other things as well. Panksepp takes a ‘hard-wired’ approach to emotions that is also categorical: the emotion of fear would be distinct from the emotion of anger.
One example of the errors this has caused is that some quite noisy voices in the ‘Pankseppian’ community think of them as ‘either/or’. What that means is you can’t, to their mind, feel fear and care together, in the way a mother might if her babies were in danger.
Another challenge of thinking of emotions in this way is that it almost excludes the existence of more cerebral emotions. It can also lead to conceptual confusion: where is anxiety as a feeling as opposed to FEAR, the flight-freeze response?
Emotions as a construct
In recent years, the opinions of Lisa Feldman Barrett have come up for discussion in the dog training world. Her popular book How Emotions Are Made and her TED talk caught the imagination of other dog trainers.
Part of the problem with this is that Barrett’s airport reading and YouTube stuff reflects her own very polarised opinions that go well beyond her research. Popular psychology editors and TED talks tend to encourage that approach. It makes good sales. In actual fact, her research is more circumspect and she too would accept some ancient physiological pathways.
Where she disagrees with people like Panksepp is that she doesn’t think emotions are hardwired. She thinks they are cognitive constructs that result from language. Researchers like Barrett accept that as we process the world, our brain reacts automatically to some of the stimuli we experience. It arouses us. That arousal has valence. Our subcortical brain categorises it into pleasant stuff or unpleasant stuff, positive affect or negative affect.
That is generally where most emotion researchers agree. We sense stuff. Then our subcortical brain appraises it simply. Good or bad.
Although she does not talk about animals in her TED talk or popular books, this view essentially excludes mammals from having (any) emotions. Since they lack language, they lack the ability to have conscious, reflective experiences.
The challenges of seeing emotion as a construct
Philosophy was the first field to discuss animal emotion in Western traditions. Well, that’s not entirely correct. Aristotle was both philosopher and scientist. However, since then the likes of Descartes have chipped in and complicated the field, and religions have had their say too.
One philosopher in particular discussed private, internal experiences only to conclude that these were so alien for any other individual to understand to be completely unintelligible. Wittgenstein used the metaphor that ‘if a lion could speak, we could not understand him’ in Philosophical Investigations written largely in the 1930s. Ironically, his metaphor about the difficulties in understanding other humans is often misunderstood and people often think it is about animals’ differences from humans.
What he was pointing out was that very human experiences are so private and personal that it is impossible for other people to understand them. Think of dyslexia, for instance. It is a very common experience for humans, yet if you are not dyslexic, you cannot imagine what it is like to be dyslexic even if you are very well studied and empathetic. As a German speaker living in an English world, Wittgenstein probably felt this more than Lisa Feldman Barrett.
Linguists have also explored the challenges of understanding the exactitude of words. Because words for emotions like schadenfreude don’t exist in English, does this mean English speakers don’t experience joy in the misery of others?
Flaws in seeing emotions as a construct
There were many philosophical and linguistic flaws in Barrett’s popular book. Wittgenstein himself settled these from a philosophical perspective in the 1930s. Linguists and semioticians have settled many discussions about how we make common meaning from words and how these differ in exactitude from one person to another.
If we need language to have ‘true’ emotions, then this excludes non-verbal humans from having emotions. Humans have probably not had language for very long, either, relatively speaking. Does this mean other Homo species did not have emotions?
It also overlooks the research of comparative psychology and ethology, focusing on animals other than humans. Given that much neuroscience research into human emotions has focused on animal models, it seems strange to then suggest that emotions are a largely linguistic, cultural concept.
Barrett’s comments are not the first time there have been disagreements between heavyweights where language is concerned. The polarisation over emotions very much mimics the disagreements of Noam Chomsky and behaviourists in the mid-twentieth century over language. Chomsky believed that language was a human universal, something innate and hardwired. Behaviourists thought it was a reinforced construct. In their aim to make a fool of Chomsky, experimental behaviourists trained a number of primates to use signs or keypads, confident that they could demonstrate language was learned and reinforced. In the ultimate scientific burn, they named one chimpanzee Nim Chimpsky.
Ironically, the behaviourists ended up proving Chomsky more right than wrong. The primates never learned language beyond function and most of the effects seen were very much subject to the Clever Hans effect, where the human inadvertently cues responses from the animal.
Other models of emotion
Seeing emotion as hardwired or learned is not the only division. There are also divisions over how many emotions may be hardwired, if they are at all, and whether they are discrete and distinct or not.
Veterinary behaviourists generally take the line that anxiety and fear are different, for example. Modern neuroscience tends to support this view. In their 2016 paper, Joseph Ledoux and Daniel Pine proposed that we see fear responses to certain threat as distinct from unpleasant feelings in response to uncertain threat. Ledoux himself does not rule out animal emotions, stating that it is impossible to know if animals have more constructed emotional experiences according to their own species and needs.
They consider emotional responses as a stimulus-response situation. We see a snake and we have a fear response. Our response to uncertain threat is anxiety.
These are different from emotional feelings, which are more cortical, complex, subjective experiences and perhaps only existing in humans. My personal understanding is that (some species of) animals can and do have more cognitive emotional feelings in their own species-experiential way. To rule out emotions like anticipation, joy, excitement, grief, frustration, disappointment and surprise does not reconcile with my life with dogs. I am more of a ‘believer’ than Ledoux, maybe because I live and work with animals.
How veterinary behaviourists view anxiety and fear
Veterinary behaviourists fall in line with Ledoux and Pine. Sagi Denenberg (2020) describes anxiety as “the anticipation of future threat” (p.207) where fear is “an emotional response to present or perceived imminent threat” and usefully reminds us that anxiety leads to the same escape/avoidance behaviours, and that flight and freeze will usually precede fight behaviours.
Lorella Notari in Horwitz and Mills (2009) says much the same: ‘fear is an emotional state that represents the response to an immediate present challenge’ whereas ‘anxiety is the emotional anticipation of an adverse event’ (p.141).
And in Carlo Siracusa’s chapter in Landsberg, Radosta and Ackerman (2023), he describes fear as ‘an emotional response due to the presence or proximity of a specific stimulus (e.g., object, noise, individual, social situation) that the pet perceives as a threat or danger’ with ‘somatic, emotional, cognitive and behavioural components’ (p.963).
In other words, there are categorical differences between fear, which is a response to current or immediate threat, and anxiety, which is an anticipatory emotional state.
Why does it matter?
This view matters in how it relates to medicine. Ledoux and Pine suggested one reason there is less evidence of efficacy with psychopharmaceuticals was because medicines affecting serotonin and/or dopamine were being used for fear responses, where they would be less effective.
We see this in recommendations from veterinary behaviourists about the use of situational medications like benzodiazepines compared to long-term medications that affect mood, like fluoxetine.
Although it would seem kind of obvious that a dog with a more diffuse response to uncertain threat would not benefit from short-term medications, there is still some lack of clarity evident in vet practice. Dogs with generalised anxiety will not benefit from situational medications to help with their daily life, and dogs who struggle with specific situations will not benefit from long-term medications that take a long time to work.
It also matters for behaviour consultants because training methods like counterconditioning or desensitisation will be less likely to work with more generalised and anticipatory feelings. It matters from both medical and a modification perspective.
Other views on emotions
Barrett’s opinions are based very much on her own research with animal models. She was unable to find specific, categorical evidence for distinct pathways, and where Ledoux and Pine suggest distinctions involving the extended amygdala for anxiety, her own work suggests it’s little more than arousal + valence. Our brains, in other words, respond only to ‘good’ or ‘bad’ and arousal alters the intensity of these feelings.
There is conflict then between emotion researchers who see emotions as distinct categories that differ in how they feel, and those who consider emotions more simply in indistinct dimensions, like Mike Mendl.
Fear for them is not distinct from anxiety other than in valence – good or bad – and arousal.
Valence is expressed in behaviour as distance-increasing behaviour compared to distance-decreasing. We move towards or approach the good stuff, and we move away from or avoid the bad stuff.
Arousal is simply the scale of wakefulness. It involves diffuse brain regions and affects the whole brain. It alters what we focus on and our mood state, our energy levels and our restfulness. When researchers who share Barrett’s views look at brain regions like the amygdala and the bed nucleus of the stria terminalis, they see that they’re not simply about fear or anxiety, but more general processes.
For that reason, they see anxiety as less aroused but more negative in feeling than fear. My own feelings on this is that it can be useful to consider arousal and valence, but anxiety can involve acute arousal that can be worse than a fear response. Also, valence is subjective and hard to measure, especially in animals.
It’s very difficult to use models like Mendl’s in application.
Implications
There is a lot of debate about emotion in psychology. It also lends to very polarised views. The truth is, like Chomsky and Skinner’s differing explanation of language acquisition, both explanations are right. However, both are wrong to insist that they are the only way to interpret things. Like language, emotions probably have constructed and subjective elements and also innate elements. Some experiences are more categorical and distinct, where others are more dimensional.
In terms of our lives with dogs, it is important to distinguish between emotions that are anticipatory like anxiety, and responses like fear. It affects the work we will do with them as well as potential use of behaviour medications.
In terms of our understanding of dogs, I do not like to take the view that they are just output machines, responding to the world around them like puppets on a string. To exclude moods and cognitive emotions as things that are not possible for dogs also leaves us open to causing emotional suffering. For instance, for many years we behaved as if fish could not experience pain. Much suffering happened as a consequence of that exclusionary view.
It’s also pretty exclusionary to view humans as exceptional. Every time people set down hard lines about the difference between animals and humans, we end up pushing back against that view. We look at dogs and know they experience disappointment and surprise. To exclude such emotions as constructs only possible by species using language and cultural transmission is to ignore what we see. And if dogs can experience disappointment, diffuse anxiety, frustration and other more complex negative emotions, then behaving as if they can not or do not could have welfare implications.
References
Denenberg, S. (Ed.). (2020). Small animal veterinary psychiatry. CABI.
Drummond, C., & Fischhoff, B. (2017). Individuals with greater science literacy and education have more polarized beliefs on controversial science topics. Proceedings of the National Academy of Sciences, 114(36), 9587-9592.
Eysenck, M. W., & Keane, M. T. (2005). Cognitive psychology: A student’s handbook. Taylor & Francis.
Horwitz, D., & Mills, D. (2009). BSAVA manual of canine and feline behavioural medicine.
Landsberg, G., Radosta, L., & Ackerman, L. (Eds.). (2023). Behavior Problems of the Dog and Cat-E-Book. Elsevier Health Sciences.
LeDoux, J. E., & Pine, D. S. (2016). Using neuroscience to help understand fear and anxiety: a two-system framework. American journal of psychiatry, 173(11), 1083-1093.
Mendl, M., Burman, O. H., & Paul, E. S. (2010). An integrative and functional framework for the study of animal emotion and mood. Proceedings of the Royal Society B: Biological Sciences, 277(1696), 2895-2904.
Panksepp, J. (2004). Affective neuroscience: The foundations of human and animal emotions. Oxford university press.
If you want more…
Feel free to check out my Substack newsletter. It’s free once a week and I do a more detailed and thoughtful paid post weekly too.
You can also sign up to my mailing list if you want details of offers or updates.