Why I don’t believe in dominance theory of dog behaviour and training

‘Dominance theory’ stands for the theory which assumes that everything in dog behaviour and training can be explained in terms of ‘dominance’, and therefore assumes the way to train your dog is basically to over-dominate it.

Scientifically this ranks as antiquated bullshit, as much bullshit as the geocentric universe model, but it’s still believed at least partially by most dog owners, and TV companies still hire charlatans to make TV shows based on it.

It is indeed a traditional majority opinion, but so was the geocentric universe model. Most people in his lifetime believed that Galileo was wrong. Tradition and majority opinion are not reliable guides to what is actually true.

Dominance theory is based on two incorrect assumptions:

  • dogs behave just like wolves;
  • wolves behave in an individually despotic hierarchical way, constantly fighting for individual control of resources.

Dogs behave significantly differently from wolves in many ways, which fit three main patterns:

  • adaptations for a scavenging lifestyle rather than a hunting lifestyle;
  • adaptations for cooperative hunting with humans — e.g. recognising finger pointing and eye-gaze direction, which wolves can’t learn to do;
  • prolonged juvenile traits — neotony, aka. ‘domestication’.

Wolves when they’re kept in badly designed zoos or conservation parks, with too many unrelated wolves or wolves from different sub-species (probably slightly different communication signals) in an area too small to establish stable family territories and thereby minimise fighting at the boundaries, do indeed behave like the traditional myths about wolves — constantly fighting for control of resources. But that is not normal wolf behaviour.

Extrapolating from wolves in such artificial and unsuitable environments is like generalising about human nature only from observations of people living in overcrowded refugee camps with people from different communities who don’t share a common language and partly because of the shape of the space cannot achieve a stable sense of security and privacy for their families.

Wolves when they live in their normal natural environment with plenty of space and the right sort of topography to be able to establish stable family territories and surrounded by mostly related families of wolves, because usually they colonise territories by pack budding and fission, rarely fight. They have sublimated sort of ritual boundary-marking behaviours, including howling and scent-marking, but as long as their territory boundaries don’t get disrupted they rarely need to fight, or not seriously. The most frequent type of interaction when you see shows of back-off (‘aggressive’) signals in stable wolf families is the parents telling the juveniles to let the younger pups eat first.

Incoming juvenile males in the dispersal stage of life are the most likely to get seriously attacked. So when you consider whether a behaviour is ‘dominance’ vs. fear-based reactive aggression from a juvenile male in the dispersal stage(ref) of life’s point of view — there’s probably no separation for a juvenile male between feeling afraid of older unfamiliar males because you are a threat to their access to females for mating and wanting to compete for breeding status yourself. There are some observable differences between reactive aggression and inter-male sexual competitive aggression (i.e. ‘dominance’), but they share the property that physiological arousal level is increased and if you react in a way that increases it further you’re likely to reinforce the behaviour rather than decrease or eliminate it. If you don’t want your dog to fight more, don’t try to terrify them into not doing it again. Don’t alpha roll them, especially not in front of a dog they’ve just had a fight with.

Alpha rolling a dog in front a dog who he’s just had a fight with makes as much sense as a training to reduce aggressive behaviours as wrestling your friend who has a phobia of spiders and head-locking him down in front of a tarantula. That’s not really likely to make him less afraid or react less to the next spider! Firmly making the dog walk on past a male dog he might otherwise react to is fine, as far as I understand, or boring them out of their reaction habit by letting them sniff around on leash within sight but far enough away from a dog they would probably otherwise react to —

The method above is just a specialised form of habituation training, which works on animals as diverse as sea slugs to humans.

Some dogs do stop reacting as before when you use dominance training methods, such as alpha rolling. This doesn’t really mean they’ve learned another way to respond or learned that the other dog isn’t a threat. It works to some extent with some dogs, but it won’t work with all dogs and it limits their learning capacity. Some dogs will resist dominance theory based training and then their humans increase their reactions, and the cycle escalates into abuse.

If you have a dog with a relatively passive temperament, you might succeed in training them into a calm submissive state or persistent learned helplessness by terrifying them with rolled up newspaper hits or alpha rolls. However you won’t then be able to train them to do much else. What a dog learns from force based training is not only the particular thing you do or don’t want them to do, but also generally that training is scary and you’re unpredictably scary.

Best case scenario with forceful training methods based on dominance theory is that you get a dog with a persistent learned helplessness syndrome, who can’t learn much else because training scares them. Worst case scenario is you get a dog who learns that early warning signals of ‘back off’ when he’s anxious or afraid will be ignored, so only the last resort ‘back off’ signal — panic biting, will work to distance him from whatever is making him anxious or afraid. Dominance theory based training is inefficient at best and dangerous at worst. It increases the risk of panic-biting, especially of children who haven’t learned dog communication yet so tend to not notice their ‘get out of my face!’ signals.

I have huskies. People often project dominance theory assumptions onto huskies even more because they look more like wolves than other dogs. In terms of neotonization — prolongation of juvenile traits, especially their lifelong learning capacities, they’re very much not like wolves and are dogs, but they are typically more independent-minded and more likely to strongly resist dominance theory based training methods. When people approach training huskies with dominance theory based assumptions, that’s usually how it escalates into abuse. I encountered these assumptions so many times and repeated this argument so many times, I’ve now decided to write it down once and for all.

my pair of ‘wolves’, ferociously competing for dominance ;)

Humans persistently like to project their ideological myths and fantasies about individually despotic dominance hierarchy being justified by natural analogies onto other species of social animals, especially dogs and horses, despite the fact that these theories have been scientifically debunked for decades. I believe that is because early childhood trauma-based authoritarian personality development is transmissible down the generations. As Fr Richard Rohr OFM says: if you don’t heal your hurt you will transmit it. I consider anyone persistently believing in dominance theory about dogs or horses, despite being shown evidence why it’s not true, a sign that they’re likely to have an authoritarian-narcissistic personality, and I consider that a disorder.

If that’s you, go to a therapist to help resolve your daddy issues, and read Theodore Adorno The Authoritarian Personality, but do not get a dog yet.

So, yes, I am judging you if you try to explain everything in dog behaviour in terms of ‘dominance’, or think that dogs do dominance behaviours to humans — which makes exactly zero sense because we cannot f*ck their females, so inter-male sexual competitive aggression between species is just nonsense.

As for the widespread and incredibly absurd idea that dogs think if you lift them up or let them sit on a higher place that they’re gaining “dominance” — this implicitly assumes that dogs process metaphors, or process information using mental representations and that they can transfer an analogous way of processing experiences from one domain to another, i.e. abstraction and metaphorical thinking, something humans can only do from about age 7, and there’s no evidence dogs have cognitive capacities beyond those of a 2 year old human. It’s also absurdly contrary to the parsimony principle of science — never assume a more complex explanation when a simpler one would fit the facts observed. The simpler explanation is that dogs feel more at risk of injury when you pick them up to avoid a fight because their bellies and vital organs are more exposed, so they react more. Same conclusion — picking dogs up to avoid a fight is not a totally great idea (although it might be the least bad option available sometimes), but different, and simpler, reasoning.

The only other species which does behave like the Mens Health magazine version of wolf behaviour is chimps. They do individually despotic male dominance hierarchy, with only occasional, fragile female coalitions limiting male violence against infants. They are like the worst side of human nature.

Some people with an anti-intellectual bias react to this explanation that they feel it is less likely to be true because it appears long and complicated. What is long and complicated about it is deconstructing the nonsense of dominance theory. Scientifically validated positive reinforcement based training is really simpler in practice and more effective. Can you get a cat to jump through a hoop and ring a bell and back again by hitting it with a newspaper? No. Can you do this through auto-shaping and clicker training? Yes, in as little as 15 minutes.

References list for this rant:

Canadian Journal of Zoology, 1999, Vol. 77, №8 : pp. 1196–1203, Alpha status, dominance, and division of labor in wolf packs, L David Mech. https://doi.org/10.1139/z99-099

Lapsed biologist retraining as a social data scientist, often writing about refugee rights advocacy and political philosophy.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store