Learning is the process by which an animal acquires a new skill. All learning is aimed at bettering said animal’s quality of life; thus it is a hedonistic process, which means that despite the existence of altruism in several species (one of which is the dog), an animal will think first of itself.
Instinct is not a form of learning. Although instinctual behaviours can be ‘mastered’ to some degree with practice, they are written in an individual’s DNA and thus cannot be greatly modified.
The following are the different types of learning that dogs employ:
Classical conditioning is the formation of an association between an external stimulus and an involuntary process. Just think of how your dog reacts when you walk into the kitchen or start grabbing your keys to leave the house. They’ve been inadvertently conditioned by you to associate a neutral action (the sound of keys) with pleasure (going outside). Your keys mean you’re leaving the house. The dog is expectant, will she be joining you for a fun walk or left alone to humour herself?
Does the name Pavlov, well, ring a bell? He was a Physiologist in the late 1800s, whose most notable contribution was demonstrating classical (also called Pavlovian) conditioning with the help of his dogs. He would show them their food, and the dogs would salivate (a physiological process inherent to digestion). The dogs would then be fed. Afterwards, he started to ring a bell as the dogs’ food was being brought out. Soon after, all it took was the sound of the bell to get the dogs to salivate.
Salivating in the presence of food is an unconditioned response, it is innate. The stimulus (food, its appearance and scent) is not neutral. Salivating at the sound of a bell ringing (a neutral stimulus) is a conditioned response, it must be learned.
While the prevalent example of classical conditioning is of salivation, an involuntary behaviour, it can also be used to condition semi-involuntary processes, such as urination and defecation. These processes are primarily involuntary with a voluntary component, just think of how we teach our companion dogs to ‘go potty’ outside.
Operant conditioning is the formation of an association between an external stimulus and a voluntary or conscious process. We see this type of conditioning when we train a dog to sit, shake a paw, come when called, etc… Any conscious (voluntary) response to a cue (a word or hand signal). The external stimulus is, for example, a command and thus the anticipation of positive reinforcement (treat, praise…) and the response is a physical and conscious reaction: sitting, shaking a paw, running towards you… Click here for more on reinforcement and punishment. It can only be considered conditioning if the stimulus increases the frequency of the behaviour or action.
This type of conditioning may be either classical or operant and works to produce a different reaction than was previously executed in the presence of a particular stimulus. A dog barks and lunges at cyclists on his walks; A savory treat touched to his nose and pulled away grab his attention, if this is repeated a sufficient number of times, eventually the cyclist will elicit a different (more desirable) reaction in the dog. Keep in mind that you need anywhere from 500-1500 repetitions to counter-condition a new response (Tout Sur La Psychologie Du Chien, 2010, Dehasse, Joël), and that what your dog finds more appealing than lunging at the cyclist will vary depending on various factors such as personal preference, the distance from the stimulus and the environment.
In theory, operant and classical conditioning happen in different contexts; in reality, they are always simultaneous. You ask your dog to sit, she does so on cue, thus you have elicited a conditioned response (the act of sitting) by using a conditioned stimulus (the verbal command, “sit”). This is operant conditioning, but she is also expectant of a treat, thus there will be salivation (sometimes even visibly so!) the most stereotypical of classically conditioned responses!
Habituation is the ability to learn to not react to certain stimuli. We see this in our city-dwelling dogs: they walk calmly through traffic and chaos, the same chaos that would terrify most ‘country dogs’.
Habituation occurs when an animal is repeatedly subjected to a certain stimulus and nothing unpleasant ever happens; he learns to not react to said stimulus. Habituation exists because emotional responses (fear, anger…) are so taxing on an animal that it is way more beneficial to be calm if it has been consistently confirmed that there is no need not to be.
Dishabituation is what happens when something unpleasant does occur that is linked to the same stimulus that the animal had been habituated to. It is the resurgence of sensitivity towards that particular stimulus. Ever heard the idiom, “Once bitten, twice shy?” This is what it refers to. Let’s say your dog is used to lying down out on the porch near the door, always undisturbed, people just step over him. One day, someone accidentally steps on his paw. The next time he’s lying there and someone goes to step over him, you can bet he’ll be wary, probably even move out of the way of his own accord. Dishabituation is essentially the undoing of habituation.
Habituation is often confused with desensitization. They are not synonymous. Habituation involves nonthreatening stimuli (seen in a dog sleeping soundly even while the TV is on). Desensitization is, instead, systematic exposure to a stimulus that produces an (undesired) emotional response, paired with counter-conditioning to change said reaction to a more desired one. A trainer who successfully redirects a dog-aggressive dog’s attention with a savory treat is counter-conditioning said dog to look to his trainer for praise (desired emotional response) rather than attack (undesired emotional response) in the presence of another dog (threatening stimulus).
Imitation is when an animal watches another animal behave in a certain way in reaction to a stimulus and shapes his own behaviour thanks to his observation of the animal, not personal experience. Although, as previously mentioned, any skill may be mastered to some degree through practice. This type of learning requires a very elevated level of cognitive function and is executed in the following manner:
3. Sequential decodification (of the cause, motivation, behaviours and consequences)
4. Motivation to behave similarly
5. Behaving similarly
(Taken from ‘Tout Sur La Psychologie Du Chien’, 2010, Dehasse, Joël).
Dogs have been shown to successfully observe and imitate behaviours, both anecdotally and experimentally. Many dog owners claim to have witnessed dogs shape their play behaviours after watching another dog in the household play a certain way. Dr. J. Dehasse observed that puppies 9-12 weeks of age who watch their mother in drug detection training are able to sniff out narcotics with more ease when they start their own training than their peers who have had no such chance to observe. Therefore, while imitation seems to be a rather rare form of learning, it is still present in dogs, albeit its implementation is not an effective method of training.
Imprinting is a type of learning that occurs during a very specific and critical period of an animal’s life.
During this time, the young puppy is exposed to certain stimuli and is very receptive to learning skills and behaviours which are not innate. Needless to say, it is very important to use this period to habituate the puppy to everything it will have to live with as an adult. The effects of imprinting are almost impossible to alter, favourable or unfavourable, good or bad, desired or undesired as they may be.
Socializing a dog to other dogs is basically impossible to erase, the dog will identify with other dogs, he will feel he is truly a dog. As a general rule, though, social imprinting becomes more easily reversible the more distant the species he’s being socialized to is from his own (canine).
It is thanks to a longer imprinting period in the domestic dog compared to most of their wild relatives (3- 12 weeks in the dog, compared to just under 3 weeks in the wolf) that has allowed man to train shepherds and the like to watch over and defend species that would normally be considered prey (sheep, goats…). Lengthy exposure to these animals while the puppy is young allows her to consider them ‘friends’ so to speak, and not food. Contrary to popular belief, it is not thanks to some genetic trait that farm dogs just know that the animals they guard aren’t to be hunted, killed and eaten. Click here for more on imprinting and behavioural development.
To be effective, a training session should encompass the following criteria:
– Training should always be incorporated in play, everything is a game. But remember: games have rules. Mindless roughhousing, for instance, is not an acceptable game.
– Nothing in life is free. You expect to be paid for work you’ve done; similarly, don’t expect your dog to do things for free out of (apparent) respect towards his ‘master’. This misconstrued, egotistic notion has no place in training. Reward your dog when he does something you like and you’ll get more of it. This works both ways! Don’t reward your dog for her mere existence, you will diminish the value of your treats this way.
– Constant reward (one repetition = one treat) to teach a behaviour, intermittent reward (several repetitions = one treat) to maintain it.
For more on how reward intervals, please check out our article on reinforcement and punishment.
Shaping is not a form of learning per se, but a method used by trainers to teach a sequence, just think of those trick dogs, when the handler says, “bang”, the dog slows her pace, limps for a few meters, then falls to the floor ‘dead’, all on one cue. This is done by slowly requiring more behaviours consecutively before rewarding. The dog must learn each behaviour separately, and then they are slowly strung together until the sequence is complete.
Capturing is another technique used by many trainers, it is useful when trying to put a behaviour on cue, especially one the dog does spontaneously. For instance, yawning is almost impossible to put on cue through traditional luring and conditioning, but if you observe your dog and mark (click and reward) every time she yawns, she will eventually learn to associate that behaviour with a reward. When this occurs, the trainer may gradually start using a cue word in anticipation of the behaviour. As you can imagine, this is a rather lengthy process, especially if the behaviour doesn’t occur often.