Cesar's Rules


 Click on the cover above to go to this book at

'Cesar's Rules' is a very interesting book. It's far from being a simple 'how to train your dog' guide. To start off with, Cesar Millan is ambiguous about the importance of training, a term he uses to refer to formal obedience training, or training in skills, like sniffer dog work.

Millan quite rightly points out that dogs may be well-trained, ie able to obey a wide range of commands, but still cause their owners serious grief. So in his view, training is secondary to fulfilling dogs' needs, especially for exercise and leadership. Millan claims to train humans and rehabilitate rather than train dogs, and in the introduction, he explains that he called in the help of professionals since he did not see his own expertise as broad enough for writing a training book.

What is training?

It's clear that Millan and his co-author have been listening to their critics. The book is much more carefully written than their previous works. Even so, a broader definition of training would be helpful, if only to remind us that every single interaction with our dogs teaches the dogs something, so training happens all the time, not just during formal training sessions.

Millan came from Mexico, where the dogs he met were often well behaved, but not formally trained. In Chapter 1, Millan describes how he worked for a time as a kennel boy at a training facility in Los Angeles. There he realised that obedience training wasn't helping many of the dogs with serious behavioural problems, so Millan has good reason to doubt the usefulness of formal training. Likewise, I live in a Spanish village where few dogs are formally trained. Few will sit on command, however. most have reasonable recall, are reliable off-leash, respect livestock, and don't mess with passing humans or passing dogs. This is more than can be said for many 'obedience trained' UK dogs.

How did these Mexican and Spanish village dogs get to be well-behaved? Before owners ever went to trainers, they trained pups in the same way as they trained kids, by spending time with them, and explaining what to do, and what not to do. The Spanish word for manners training is 'educar', or 'to bring up', used for both dogs and kids. Informal training is just teaching dogs rules, such as 'it's OK to chase rabbits, but not chickens'. As Ian Dunbar notes in 'Cesar's Rules', commands, such as sit, or recall, help impose those rules. Often we don't realise how much training we do through everyday interactions which tell dogs what our rules are, and help explain to the dog how the world behaves, for example, on a walk 'you don't mess with cars, you stay near me, and the cars don't mess with you'.

Dogs have more freedom to be off-leash in rural areas, and could be off-leash much more in 'the old days', when there were fewer cars and fewer legal restrictions. Most owners in developed countries now face more difficult conditions, because streets are more dangerous and there are more leash laws. We're often out working, and come back tired, to find a hyped-up dog desperate for a run. Many of us turn to professionals for help.

Among dog professionals, there's a separation between 'trainers', who teach owners how to get their dogs to behave well, or learn skills, and 'behaviourists', who specialise in problems between owners and dogs. ('Behaviourist', confusingly, can mean either a specialist in canine behaviour problems, and a follower of a school of psychology called 'behaviourism'.) There are 'trainers' who see their job as simply instructing owners to teach dogs to obey a few commands, but there are many dedicated trainers who help owners to use commands to teach their dogs rules. Good trainers look at messages we may unwittingly send our dogs, at what we may unwittingly be training our dogs to do. They also look at non-training issues, like exercise, diet, and the environment the dog lives in. In practice, there's little difference between a good trainer and a good behaviourist. We're training our dogs all the time, and professional trainers can help us to improve our dogs' behaviour. Millan is right to say that teaching commands alone doesn't make for a well-behaved dog, but he underestimates what a good trainer can do with those commands.

This is not a 'how to' book!

Why do we train dogs? Mainly because we want them to behave in certain ways, and respect certain rules. There's a very interesting summary of what owners want from their dogs in Chapter 2. It's presented with little discussion, mainly to point out that owners tend not to want much, and what most of us do want is quite easy to achieve. I feel there's a lost opportunity here, because the survey could be a useful starting point for the 'how to' part of 'Cesar's Rules'.

Top of the list of what owners want is recall. Here the survey could be more strongly linked to advice later in the book on teaching dogs to come back when called. Then comes getting on with other dogs and people, especially kids. Again, this could be more strongly linked to Millan's advice on dogs and kids. Millan's insights into the abilities of well-behaved, confident dogs to teach other dogs are also relevant. What else do owners want? The next highest wish in the survey is that we want dogs to refrain from trashing the house, and/or howling when we're out. This point could be more strongly linked to Millan's emphasis on fulfilling dogs' needs for exercise. Some dogs do feel acute anxiety when separated from their owners, but many dogs benefit from a run before being left alone, and they're more likely to sleep after a run.

Linking the Chapter 2 survey more strongly with later advice would have given 'Cesar's Rules' much more coherence as a 'how to' book. but this book is more than a guide on how to train, it's also a discussion of different approaches to training, something of a change in direction, and a defence of Millan's own approach in response to some quite vociferous criticism.

The mature Millan

Millan was initially criticised for being overly confrontational, and for putting too much emphasis on dominance and punishment. The young Cesar could easily fit into the Spanish village where I live. Men here feel that they should show physical courage, let a dog know who's boss, and they tell you that dogs and kids need discipline. (Men need discipline, their wives say!) Confrontation makes for dramatic TV, but that isn't the only reason why Millan became very popular very fast, he was also addressing a need. Much of what he said was simple common sense that some Californians had forgotten, dogs can suffer from 'cabin fever' if they aren't walked, and if you don't spell out what the house rules are, dogs will make their own rules.

'Cesar's Rules' reflects a more mature Millan. He has less need to prove he has balls, physical courage. He recognises that most owners don't have his skills for 'reading' dogs, or his timing, so they may end up being bitten if they try confrontational methods that he can get away with. He also recognises that confrontational methods can cause unnecessary stress for dogs. Whether or not to confront is an art, rather than a science. The question is 'will tackling this head-on solve the problem fast, or will it make it worse?', and the answer really only comes from experience, knowing what you feel inside, and reading the dog. Millan's recommendations in 'Cesar's Rules' are for a much less confrontational approach than that shown in his early TV programmes. He has also moved on from his earlier crude model of 'dogs as wolves', of which more later.

Millan explores behaviourism

A great strength of this book is its exploration of behaviourism in Chapter 3. Behaviourism is an approach from psychology that links learning and behaviour to rewards and punishments, and is the orthodox approach to training in the UK and US. Millan and his co-worker's previous mentions of behaviourism were a reaction to people who call themselves 'pure positive' trainers, by which they meant using only positive reinforcement. However, 'positive' in behaviourist language just means 'adding something', either adding a reward to reinforce a desired behaviour, or adding a punishment, to decrease undesirable behaviour. Someone who believes strongly in administering punishment could thus logically describe themselves as 'pure positive', so it's a confusing term. Furthermore, while all trainers would agree that harsh punishments can undermine trust, it's not clear that dogs can be trained to follow household rules without 'adding something that stops a behaviour'.

A key weaknesses of behaviourism is that it isn't clear whether 'punishment' is a stimulus that a dog doesn't like, or whether, as Ian Dunbar puts it in Chapter 3 of 'Cesar's Rules', just something that makes a dog less likely to behave a certain way, without necessarily being unpleasant for the dog. Now, if you have a good relationship with a dog, and say 'please don't do that' (eg chew the chair leg), chances are the dog will listen and obey, and eventually stop chewing the leg (especially if you offer a chew-toy as a permitted chewable object in place of the chair leg). A good relationship means trust, and wanting to know what the rules are, because rules can make the world a safer and more enjoyable place to live in.

In good relationships, most dogs and people can handle the message 'I love you but'. Millan himself illustrates this when he tells owners that they have screwed up with the best of intentions and need to change their ways. He builds trust with his message that the owner is a good person who wants the best for their dog, and then adds something - criticism - that changes the owner's behaviour. Millan has a talent for developing trusting relationships fast. One reason for the ambiguity in the behaviourist definition of 'punishment' is that there's a lot more going on between dogs and people with a good relationship than behaviourism can adequately explain. Ian Dunbar hints at this when he argues that punishment doesn't have to be aversive, because humans training dogs are in a different position from computers shocking rats (Chapter 3 p101). Studying behaviourism can help us realise when we are rewarding bad behaviour without realising it. But it isn't a lot of use unless supplemented by an understanding of human-dog relationships, and, as Millan notes at the end of Chapter 3, behaviourism ignores canine inbuilt tendencies.

Millan's last sub-heading in Chapter 3 is 'Dog Training vs Dog Psychology'. This subheading implies that training is the application of behaviourism, and is possible without understanding what makes dogs tick. Yet training dogs to follow rules has to take into account the nature of dogs, and their relationships with their owners. Defining training as the application of behaviourism, as Millan does, gives you a narrow view of training, confined to responses to external stimuli when dogs learn simple tasks. But training dogs to follow house rules involves understanding what's going on inside the dog, its feelings about what is going on in a relationship, or inbuilt desires to do some things, desires which can override any rewards that owners might offer.

Some people learn about the nature of dogs just by being with them a lot. In the academic world, ethologists have systematically studied canine inbuilt tendencies. Ethology was popularised among 'dog people' by Bruce Fogle. whose 'Dog's Mind' was very influential in the early 1990s.

The nature of dogs, and how they learn from each other

Millan was initially a great fan of Bruce Fogle, who popularised the idea of using wolves as models for understanding dogs. We may put Fluffy in a little dress, but she's essentially a designer wolf. The young Millan found Fogle's view of dogs as wolves especially appealing. Fogle's message was that wolves have strong hierarchies, so humans need to show dominance in order to control dogs, and this fitted with the young Millan's view of dogs. Since Fogle's work was published, a lot more has been learnt about wolves, and ways in which dogs differ from wolves, for example, the work of Adam Miklosi's team in Hungary. Miklosi's team has raised wolves, and compared their development to that of dogs. They've spelled out important ways in which dogs differ from wolves. For a long time, Millan seemed deaf to criticism that his ethology was flawed, but this is starting to change in encouraging ways.

The mature Millan is much more cautious about using the term 'dominance', and he seldom uses the term in 'Cesar's Rules'. He has moved on from Fogle, but is endearingly loyal to his early hero, whom he mentions at the end of Chapter 3. Of course all of us are likely to be seen as 'old fashioned' by later generations. Much of human knowledge is cumulative, we build on the backs of those who have gone before us. Fogle helped owners to understand that their pooches were not little humans, but in some ways were like wolves. His arguments were often flawed, but there was still some truth in them. Now that Millan has moved on, it would be good to see him learn from more recent research, and think a little more about what he calls 'the power of the pack'.

Millan often notes how dogs can influence each other for the better, and there are several mentions of this in 'Cesar's Rules'. Well-behaved, confident dogs can help other dogs to become well-behaved and confident. Sometimes dogs can teach each other faster than we can teach dogs - for example the story of Daddy and Viper in 'Cesar's Rules'. This learning is wonderful to watch, and it's a strength of Millan's that he's able to understand and use this canine ability. However, calling it 'the power of the pack' is perhaps misleading. Dogs don't have packs in the same way that wolves do, but are far more flexible. They can learn to get on with unrelated dogs they live with, and a range of other dogs in different situations. Dogs from different households that are often walked together may form a sort of 'pack', in the sense that they act as a group, for example, sniff grass together. Not every group of dogs is a 'pack' in any meaningful sense, even if they live in the same house. Even if they're like a 'pack' in the sense that they coordinate their activities and get on well with one another, 'packs' aren't always benign. Dogs can teach each other bad habits too, like barking at bikes!

Teaching skills to dogs

Dogs with particular skills appear throughout the book, such as Viper, the highly trained and neurotic cell-phone detection dog. Millan also describes the work of his early heroes, Hollywood trainers, in Chapter 5. He looks at 'herding', hunting, protection and cancer sniffing dogs in Chapter 8. He's aware that some dogs, of a type a friend once described as 'busy', need interesting things to do, or they tend to get up to mischief. Millan has helped to publicise and promote skills training, but perhaps both he and most pet dog owners underrate the enormous benefits that most owners could enjoy from teaching their dogs skills. Chapter 4 sets out 'Cesar's rules for a teachable dog', and there's no mention in these rules that dogs need jobs. The survey in Chapter 2 shows that most owners surveyed place a low value on skills training. Maybe Millan focuses on fire-fighting rather than skills training because that's what people ask him to do.

Millan quite rightly notes that some people divorce skills training from the needs of the dog as a whole. However, skills training can help dogs and humans learn to communicate better, trust each other more, and take more enjoyment in each other's company, so it can go a long way to solving behavioural problems. In particular, teaching retrieving includes teaching dogs to drop precious objects on command. It's a useful way of preventing and tackling 'resource guarding', and of course can be helpful if you want an object retrieved that you can't get to and the dog can. Skills training is where a real partnership begins between dog and owner. The dogs are doing activities that are natural to them, and are rewarding in themselves, such as using their noses.

Millan has adopted a term from Ian Dunbar, 'life reward', to describe a task that is a reward in itself. Though Dunbar and Millan have different approaches, they found they had a lot in common when they met. Like anybody passionate about a particular field of study, Millan is happy to learn from anyone who has something useful to tell him.

Cesar Millan and Ian Dunbar spend a couple of days together

Millan's meeting with Ian Dunbar is described in Chapter 6, which focuses on Dunbar's work, and his approach to off-leash training. This is a curious chapter. Millan is suitably deferential to a high-ranking male, mentioning Dunbar's qualifications, achievements, attractive wife, and genial personality, and he presents Dunbar's advice with no comments. Dunbar's 'seven rules' include advice to select an eight-week puppy that's "housetrained, chew-toy trained, has been taught to come, sit, lie down and roll over on cue, and has been socialized and handled by at least 100 people." (p191) However, any pup whose owner subjected him to that regime before eight weeks could end up a nervous wreck! You can start housetraining a small pup, but dogs don't develop full bladder control until after eight weeks. Yes, it's useful for pups to meet children, men and women when they are small, to get used to different voices and ways of behaving, but being handled by 100 people isn't necessary, and indeed would probably be very stressful. 

Dunbar's advice "always have your visitors offer your puppy/dog a treat" (again in the 'seven rules') is also impractical. Firstly, some visitors don't want that sort of close contact with dogs, and there's no reason why they should have to tolerate it, especially if dogs make them nervous. Secondly, having visitors always fuss and treat dogs can encourage them to become overexcited when visitors arrive. My mother's advice to her many visitors was much more sensible 'Just ignore the dogs'. Yes, they did occasionally sneak a cuddle from a particularly dog-friendly visitor, but when both dog and visitor had decided they liked each other.

'Hands off' or 'hands on' training?

The whole of Chapter 6 focuses on 'hands-off' methods, which have tended to dominate professional training in the US and UK. As Dunbar explains in this chapter, the idea is that rough handling can undermine trust, so it is better to have no handling when teaching commands. "Most human hands cannot be trusted. Of all the humans who can't be trusted with their hands, I guess men and children are probably the worst. They do a lot to dogs that that spook dogs out" (p162). Dunbar uses treats and his voice, eventually phasing out the treats. Millan himself mentions at the end of this chapter that he prefers touch as a more natural way of teaching dogs. Dunbar is right in that many people do spook dogs by the way they touch them, but surely the solution is to teach people how to touch dogs.

Trainers and dog behaviour specialists tend to see novices, or people with difficult dogs rather than people who've successfully brought up dogs to obey rules. This can give trainers a bleak view of pet owners, and maybe Dunbar underestimates what most owners are capable of. Most people can understand and follow basic advice if they realise that it's very important.

There's a big disadvantage to 'hands off' methods, which is that dogs need to accept being handled if they're going to get veterinary attention, and be reasonably safe around people.Pups trained with 'hands-free' methods may not learn to trust hands. Gentle, firm handling which is sensitive to the dog's reactions can help to foster trust, whether it's part of formal training, or just part of everday interactions. There's also no point handling dogs appropriately when they're being formally trained if they're handled inappropriately at other times. All handling involves training. Hands-free plus voice methods are also less useful with deaf dogs than are touch and gestures. Most owners don't start out with deaf dogs, but dogs often go deaf as they get older. 

Millan, then is outside the orthodoxy, in his use of touch. His experience as a groomer taught him a lot about how to touch dogs, and he could say more about handling dogs in ways that foster trust. How pups are handled affects how they respond to touch as adults. Much of the skill of knowing how and when to touch comes from 'reading' dogs. These are topics that the mature Millan could fruitfully explore, now he's moved on from his 'lion taming' youth.

Read this book!

Like any book, 'Cesar's Rules' isn't perfect. It's not a simple 'how to' training book, though it contains plenty of useful training advice. Millan is a 'natural trainer' who has learnt by trial and error. What he does comes from a gut feeling developed from experience, so it 's sometimes difficult for him to describe how he trains, especially as he defines training narrowly, and does not even see himself as a trainer. More than a 'how to' book, this book is also an exploration of different views, and a pooling of different talents. It's a book that will activate a few brain cells, force you to think and look at your dog, which is a pleasant change in the world of dog books.

Alison Lever

Thank you to Diana Attwood, Wendy Hanson, Tiffany Jho and Janeen McMurtrie for helpful discussions and arguments about this book. Thanks also to Berit Aherne, a 'real trainer' who first taught me about what dogs are capable of.

Further Reading

Don't Shoot the Dog: The New Art of Teaching and Training by Karen Pryor, popularised operant conditioning among dog trainers. However, Mary Burch and Jon Bailey' How Dogs Learn gives a clearer account of the way operant conditioning is used in dog training. Despite the title of Burch and Bailey's book, operant conditioning is not the only way to understand how dogs learn. It can be a useful explanation, but has limitations. For a fuller account of how dogs think, and up-to-date research on dogs and wolves, Adam Miklosi's Dog Behaviour, Evolution and Cognition is very thorough.