What is the difference between how dogs learn and how dogs think?

What is the difference between how dogs learn and how dogs think?

Kevin Behan says that dog’s don’t think, they are rather, a “feeling” an “emotion” – and in simple terms they work like an electrical current, conducting everything they take in from those around them, aligning themselves accordingly and working totally as forces of energy. Cesar Millan, although in different ways to Behan, also says that to dogs, we are “only energy” and that they reflect everything about us. In the words of Behan, this super conductivity and being of emotion means that the dog “becomes us” and blends with us into one “group body”. They are no longer a separate entity and thus – what we think – is what they reflect. It is a very interesting concept and having witnessed many similar examples to those he gives in his book for myself – in both horses and dogs – I have to say that I believe there is something in it. I might not explain it in quite the same way that Behan does, but there is more than a grain of truth in what he says.

Dog food puzzleA good example which helps us start to understand the the difference between the way dogs learn and the way they think is that dogs don’t generalise. This can be the downfall of some trainers who might not make it clear to an owner that just because the dog is excellent in the training hall, it might not perform the same way out in the field, at home or on a visit to a friend’s farm. Dogs think very much in the moment and in the place where they are. They associate things exactly as they happen and so it is up to us to help them learn to make those associations in different places, by many repetitions in many different locations. For example a dog who understands “sit” very well at home, might not be able to offer the same sit when out in the countryside, or if the person is on top of a horse or at a distance. To the dog things are different. The dog does not think it through, it just responds in that moment and place in time. Dogs will learn to associate a “word” like sit with many places and many situations if we help them and this is best done by reinforcement and this is the way dogs learn that what they try is or isn’t correct. That reinforcement is best done positively using things that motivate dogs – treats, praise, play or as one horsemanship trainer says – “praise, recognition and material things”!

Dogs, and in fact most animals, learn primarily by trial and error. When we use this in training we call it shaping – and it works the same way as the children’s game of “yes/no” or “hotter/colder”. The correct answer gets a yes or a hotter response, the wrong one, a no or colder response. In training dogs and thus helping them learn what we want, we can increase the chances of a behaviour occurring by reinforcing the correct or desired behaviour. When the dog uses trial and error – if it gets it wrong the behaviour is either not rewarded, punished or ignored. The dog learns “error = don’t repeat that behaviour”. When the dog gets it right and we reward the desired behaviour with reinforcers such as praise or food, the dog learns “correct try = repeat the behaviour”. Over time the reinforcer, command and behaviour become associated and so the dog can repeat the behaviour on command.

In the learning process, reinforcements fall into two main categories – primary and secondary. Primary reinforcers are those related to biology – such as food, drink, sex and some kinds of touch. Secondary reinforcers are those related to social conditions such as praise, smiles, attention, clapping, toys, patting/scratches etc. Dogs have to learn that praise means something positive and so secondary reinforcers become reinforcing by being paired with primary reinforcers.

The best learning therefore is through reward and motivation. Stimulation can be play or something that encourages a breed characteristic – and knowing a breed’s characteristics (or in a mixed breed, the characteristic of the most dominant breed in that dog) initial learning can be improved upon. For example, a dog that is motivated by smell (and most dogs will use their nose well hence why Cesar Millan likes to get the dog motivated and understanding via its nose) will respond well to maybe searching things out – like buried toys, tracking and so on. Millan for example would use a treat to help a dog learn to be confident on a treadmill as it will follow the scent in order to move forward as the treadmill turns. A dog that is motivated by sight (collies that use the eye stalk for example) will really enjoy watching – eg eye on the ball to fetch it or chase it and often do well in tasks like agility and obedience where they really focus on their handler. Often the job and the task in hand are reward enough for the dog once they initially understand what to do. As Vilmos Csányi (the Hungarian ethologist who wrote “If dogs could talk”) says – a dog considers humans it’s social group and so working in that group and being recognised by that group is motivation enough for most dogs.

Dogs and in fact all animals, learn far better through motivation and positive training. It is far better to find the positive and redirect unwanted behaviour than to always punish what we don’t want. There should always be far more positive reinforcement in training than anything else – as although trial and error will (even in the wild) result in punishment being delivered (eg approach snake, get hissed at/spat at/bitten – don’t do it again!)… these punishers are not unreasonable or delivered with emotion. All animals will almost certainly shut down or become dull and unwilling partners (at best) or aggressive and dangerous (at worst) with too much punishment. Nothing is worse than an animal that is in a state of learned helplessness – something that is only created by humans.

dog thinkingDogs probably learn and have memories more linked to their actions than events, because it relates to the time they were hunters. For example, there is little purpose in hanging around for prey where it just disappeared as it wastes time and effort and does not offer reward (an event). It is better to move on to a new hunting site and pick up a new prey as soon as possible. But, for example, in the wild, a fox that has buried food (an action of its own) is likely to have a long term memory of it and will remember where to find it. Dogs therefore seem to think in terms of actions and consequences and in terms of the permanence of objects – such as having the ability to memorise landmarks. They seem to be able to construct complex mental maps and can work out short cuts for themselves.

Dogs also learn by stimulus enhancement – that is, they keep an eye on what other animals are up to in case it sees something useful. Dogs in this way also learn from each other and this transfer of skills is something that helps dogs to learn quickly rather than each one having to learn everything from scratch by itself.

When it comes to thinking – dogs think much more in the moment and they may think by what they feel/smell/see at that moment. They do associate but they don’t plot, plan or think in the way humans do. The way they think links with the way they learn – through more instinctual needs and through energy. As dogs are an emotional reflection of their owners, one might say that the dog is, in many ways, a manifestation of human thinking! Dogs certainly do seem to reflect their human’s emotions and this can effect their learning, the reason why so many get unbalanced with their emotion/fear/anxiety etc getting in the way of their learning. Just as with humans, if we are not relaxed, enjoying ourselves or having fun, we find it much harder to learn and what we do learn is less easily retained.

Csányi points out that sleeping dogs can be seen running or hunting – and points out that we can not deny the validity of emotions in an animal that dreams – and having emotions and being an emotional being, links to thought. He interestingly discusses how humans have selected for breeding, dogs which “reflect what we find human” and so at his conclusion he wonders how much further this will take us in levels of communication as we select dogs increasingly to have some insight into our own human thought. There is some suggestion that dogs may have or be heading toward, some kind of self-awareness. As he says, dogs experience pain, anger, fear, joy, sorrow and excitement and they keep track of every morsel and every caress. They want to be a part of everything and will know who is the person most in tune with them and who “likes them best”. These are processes of thought.

Csányi also says that dogs, in contrast to all other animals (other than humans) are capable of empathy – “emotional identification in dogs is not only the attainment of some kind of passive harmony, but enables them to carry out intelligent actions”. He also states that “in contrast to all other animals, dogs are capable of following rules”. In his chapter “Dogs Understand a Lot” he gives a number of stories to illustrate: “that dogs are quite advanced in their analytical abilities, because (the stories show) they can not be explained in simple terms of associative learning” and that because of this “we probably need to assume more complex mental processes”.

John Bradshaw raises an interesting question regarding the cognitive abilities of dogs as he says that they are likely to be “qualitatively different from ours because they have been selected to match the canid lifestyle”. He says that the usefulness of guide dogs, for example, depends on their ability to “think outside the box” and “to use their canid brain to predict what is going to happen next in the ever-changing environment with which their owners are interacting” – he says that this skill is likely to have come from the wild canids’ ability to predict their prey’s next move. He goes on to say that finally, after years of avoiding the topic, both biologists and psychologists are addressing the subject of how dogs “think” and looking at the more complex things that their brains can do. He bears out what Csányi says: that what is becoming more clear “is how domestication may have affected the dog’s ‘intelligence’, and also why it seems to mesh so well with our own”.

HelperDomestic dogs, apparently, can outperform even chimpanzees in some very specific ways and although chimps overall intelligence may be greater than that of dogs (depending of course on how we define that!), it seems that dogs have a “special brand of intelligence, unique in the animal kingdom, which they ‘co-evolved’ with humans as part of the process of domestication”. One area in which dogs outperform chimps and which is part of this process of domestication is their ability to extract information from what humans are doing and in particular in their ability to read human faces and gestures. It is important to remember here also that if we provide an interesting and stimulating environment for a puppy it will develop a larger brain, more cells, bigger cells and more interconnections between them. Thus we directly influence the development of the dog’s mind – both in the way he will think and in the way he will learn. It is in this way that it is most likely that humans and domestication have developed the dog’s ability to learn and to think beyond his original pre-domesticated capabilities.

I think it might be useful to quote directly from Bradshaw’s book at this point (In Defence of Dogs) because it can’t really be put any better:

“Analysing canine intelligence is not straightforward. Just as we can never be sure precisely what the inner world of canine emotion is really like, we shall probably never be certain whether dogs think the same way we do. Science has so far been unable to tell us how self-aware dogs are, much less whether they have anything like our conscious thoughts. This is not surprising, since neither scientists nor philosophers can agree about what human consciousness consists of, let alone that of animals. However, it is possible to examine scientifically whether dogs can do various things, and then to infer the kinds of thoughts they might have, bearing in mind that, as dogs, they may not have the same priorities that we (or other animals) have in the same situation”.

He warns us that we must not treat dogs as little people rather than dogs, because when we do this our actions become misleading and the dog becomes confused and distressed. This is why unfortunately, more and more owners find themselves requiring the services of behaviourists and turning to people like Cesar Millan for help.

To Conclude:

In a recent newsletter sent out be Cesar Millan he answers the question about the difference between dog training and dog behaviour in his own way. The excerpt quoted was taken from his book “Cesar’s Rules” and he puts it in a very easy to understand way (see below). It also brings us back to the difference between the way dogs learn and the way they think. Learning is mostly by operant conditioning (voluntary behaviour elicited by various means) and operant behaviours are controlled by consequences (which can go under a broad umbrella of trial and error learning). As Bailey says, in dog training these consequences are mediated by the trainer.

Bradshaw states that the simpler forms of learning enable the dog to piece the world together and this is what allows us to train them. He says that they also have the ability to think for themselves and have a knowledge of the world around them and other animals in it – obviously including humans.

Thinking and learning are linked but not the same. This link can broadly be seen as dogs linking, or making connections between actions and consequences. Fogle explains another link between thinking and learning in his book “The Dog’s Mind” in the chapter on the brain: “It is the temporal lobe’s connection to the hippocampus and the hypothalamus that might explain the relationship between memory and the emotion INTEREST. All good dog trainers know that a dog must be interested in training otherwise the exercise is hopeless.”

Enabling a dog to think and “be” will help it learn. But learning will not necessarily help it think or “be”. As Csányi says: “Consistency is the foundation of the canine mind” and so if we are inconsistent when helping a dog learn – it can’t get the right answers from trial and error learning, because they will differ. This leads to confusion, worry, fear and anxiety. Reinforcement is needed consistently for the right thing so that the dog knows it is the right thing. He also explains that the dog, despite the human, is often able to interpret (and excel in) quite complex situations and can act appropriately based on those interpretations. Dogs excel in social understanding and interpretation of human expression and emotion. This indicates that the thinking of the dog is through energy, as emotion and actions are all forms of such, just as Kevin Behan details in his book. As he says:

“A dog’s thinking is about creating balance. What they think in a moment is picking up and reflecting energy. In many ways our dog’s thoughts and deeds are ours. What matters most to our dogs is what we feel and the emotion that’s invested in the things we do, say or think. They study our every move and read us”. Behan says they don’t have thoughts, they just “are”.

To end with the quote from Cesar Millan (referred to earlier) it brings us back to the difference between behaviour and training and in fact also to the difference between learning and thinking. It answers both questions in such a simple and succinct way!

“Training vs. Balance: What’s the Difference?

There is so much confusion about dog training out there—what works, what doesn’t, what’s best, what does it even mean? And for me, training is about conditioning. What I do is not “training” in the traditional sense, as what I care most about is “balance.” That means fulfilling the instinctual needs of the dog. And when a dog is balanced (thinking), training (learning) is much easier to perform.

(Note: the bracketed red words are my own and do not appear in the original article!)