Instincts, reflexes, associative and non-associative learning and other types of learning.
The ‘ Learning is experiential and behavioral change is key to ensuring the survival of an individual, as each animal, to survive, must be able to adapt to the changes to which it is subject in its environment. There are different types of learning, from more “automatic” forms, as in the case of behaviors caused by psychophysical alterations (ie reflexes and instincts), to more complex ones.
Reflexes
Reflexes are behaviors that occur following specific stimuli and are characterized by rapidity , automatism and rigidity . They are therefore spontaneous, physiologically determined reactions. Examples are salivation and the patellar reflex, ie the leg extension response following pressure below the knee [1] .
Instincts
Like reflexes, they are behaviors deriving from specific stimuli and are stereotyped , ie they occur according to sequences of movements that are always similar. They too are genetically determined and not controllable. Unlike reflexes, however, instincts can be more or less flexible [1] . Examples of this are the calls of the parents from the little ones and the courtship dance of the manachin ( Machaeropterus deliciosus ).
Learning
The real learning occurs when the learned behaviors are very flexible. Consequently, they allow you to better respond to situations with respect to reflexes and instincts [1] . Not surprisingly, therefore, all animals are likely to be capable of learning from their own experiences. The types of learning are many. The most important and widespread are the following.
Non-associative learning
Following the repeated presentation of a stimulus , an animal is able to learn that it continually presents itself. As a consequence of this it can react with two learning mechanisms: addiction or sensitization . The first case is faced with insignificant stimuli: the magnitude of the response to a stimulus decreases because it is a useless waste of energy . In the second case, however, there is an increase in response, due to potentially harmful stimuli: in this way the effectiveness of the response increases., which can be for example the avoidance of the noxious stimulus. For example, the first encounter with a predator may result in an increased response in subsequent encounters [1] .
An example of learning by habituation can be observed in the sea hare (Aplysia californica), a mollusk, by stimulating the siphon in a tactile way, for example through a spray of water. The first time a gill reflex will be observed, but the more times the stimulus is repeated the less the reaction will be strong, as it is a stimulus that does not follow a negative effect and is therefore perceived as harmless. At the nervous level it is observed that the sensory neurons of the siphon produce less and less neurotransmitters directed to the motor neurons [1] .
In sensitization , on the other hand, there is a greater release of neurotransmitters. An example is the escape response to the predator : in subsequent encounters with this, the escape will be faster or more ready. This type of non-associative learning is also called pseudo-conditioning because as an effect it resembles classical conditioning: the animal learns that a response, in this case negative, follows a stimulus. However, the mechanism is different, because in this case there is an increase in physiological reactivity, while in conditioning there is a real association between stimuli and behaviors [1] .
Associative learning
In associative learning, animals learn which environmental events herald or cause biologically relevant events so that they can respond appropriately to them. For example, if a certain sound announces the arrival of a predator it is evolutionarily advantageous to develop the ability to associate that sound with that predator and flee in advance of the predator’s arrival, as in the case of moths with bats . Associative learning can be classical (or Pavlovian ) or operant (or Skinnerian ) [1] .
Classic associative learning
In classical associative learning, an animal is repeatedly subjected to a neutral and unexpected stimulus (event E1) and then a significant stimulus (event E2), which normally triggers an innate response in the individual (unconditional response). At a certain point the individual will learn to connect the two events, also associating the same response triggered by the significant stimulus to the neutral stimulus. The latter is called a conditional response [1] .
The first to try his hand at this study was the physiologist Ivan Pavlov, who at the end of the nineteenth century carried out an experiment on dogs. By measuring salivation, he observed that salivation increased (unconditional response) when the animals saw the food (event E2). On the other hand, nothing happened when a bell rang (event E1). He then tried to play it every time he presented them with baby food. After dozens of sessions, dogs showed salivation even when they heard only the bell, without being given food (salivation had therefore become a conditioned response) [1, 2] .
What happened? To a previously neutral stimulus, to which the animal therefore did not present a physiological or behavioral reaction, the individual had associated an event with a meaning (to which he therefore normally reacted with a response). That is, he had developed a conditioned reflex towards a stimulus that was previously indifferent to him. In order for the animal to develop a conditioned response, it is also important that the neutral stimulus is unexpected [1, 2] .
Associative learning is not always possible
First of all, there is no conditioning when too much time passes between the neutral stimulus and the unconditioned one. The mental association between two stimuli almost always occurs only when the interval between them does not exceed a few seconds . An exception is that of food and nausea: a mouse that takes a new food and subsequently feels a sensation of nausea easily associates the two episodes even if they occur seven hours apart [3] .
Second, conditioned responses must have biological meanings . For example, rats cannot learn that after a certain sound they will experience nausea or that after a specific taste they will suffer electric shocks, probably because in nature it is impossible for a sound to cause discomfort or for a food to cause pain in the paws [3] .
It is also generally not possible to condition animals if multiple conditioning stimuli occur. If an individual is presented with two consecutive neutral stimuli, before reinforcement, he will learn only the first stimulus and not the second. This phenomenon is called blocking : the second stimulus alone will not be able to trigger the conditioned response because it is not an unexpected event [1] .
A further case is that of overshadowing , a phenomenon whereby if the individual is subjected to two stimuli it will respond only to the more intense one, because the less intense one will be “overshadowed” by the other. For example, if we add the variable “I turn on the light” to Pavlov’s experiment, together with the sound of the bell, when we bring food to the dogs, they will probably associate salivation with turning on the light and not with sound, because light is a stimulus stronger than sound [1, 3] .
Operative associative learning
In operant learning, also called by trial and error, the individual learns to associate a voluntary action with positive or negative consequences [3] . For example, a mouse has to push a lever in order to reach the food: it will be up to him to learn that behavior to fulfill his desire. It will initially engage in a series of random behaviors; then he will learn to carry out a certain action by dint of seeing that the administration of food follows [1] .
The functioning of this type of learning is based on the law of effect , enunciated by Edward L. Thorndike in 1989. According to this law, an individual is more likely to repeat behaviors accompanied or followed by contentment and less likely to repeat behaviors accompanied or followed. from being unwell in a certain situation, when this situation is repeated [4] .
To influence the behavior of animals in operant conditioning reinforcements and punishments are used :
- food is a typical example of positive reinforcement , which stimulates the animal to repeat a given behavior because it associates something pleasant with it. A sort of “reward”; another example is praise, which serves as a positive reinforcement in horses [5] and dogs.
- The negative reinforcement occurs when, as a result of a behavior, it subtracts an unpleasant stimulus to the animal. In this way the individual is stimulated to carry out that behavior [5] . For example, a child will stop being scolded if he tidies up the room.
- The positive punishment consist in imposing a negative stimulus to the animal as a result of an undesirable behavior; the individual, therefore, will associate unpleasant consequences with such behavior and stop implementing it. This is the case with physical violence and electric shocks. Positive punishments are often not effective because they are not applied correctly, too frequently and disproportionately to the behavior you want to inhibit. Thus, the animal can get used to it or fall into depression without changing its behavior [5] .
- The negative punishment you have when you deny something pleasant to the animal. Again it will reduce the associated behavior [5] . An example of negative punishment is to ignore your dog when he jumps on you: it is unpleasant for him not to have your attention, so in the long term he will learn that he does not have to take that action.
This type of learning was discovered by the psychologist Burrhus Frederic Skinner in the 1930s , thanks to cages that bear his name. Skinner’s boxes are cages in which it is possible to insert small animals to carry out observation experiments and ethological experimentation. Inside them there may be objects with which individuals must interact; following this interaction they can receive positive or negative stimuli [1, 3] .
Examples of application of Skinnerian conditioning
Operant conditioning has been used extensively since its discovery: the most diverse animals (including parrots, reindeer, raccoons and whales) have been conditioned for fairs, zoo exhibits, television broadcasts and commercials [5] . Thanks to operant conditioning it is also possible to make an animal perform several actions in a row or change its behavior.
In fact, shaping (or modeling) consists in reinforcing the animal’s behaviors since they are only a sketch of what you want and then, gradually, reinforcing the behavior so that it progressively approaches what you want [1] . For example, if you want to make a dog perform a sequence of actions , such as “paw” – “ground” – “sit” – “roll”, you must first teach him the first action, giving him a treat when he correctly executes the command, then the second, giving him the prize only after he has made both “paw” and “earth”, and so on.
Cases in which operant conditioning is not possible
But even this kind of conditioned learning isn’t always possible. When the action you want to teach the animal is in contrast with strong instinctive behaviors for its species it is possible that there is a shift from the conditioned response towards the instinctive response. A rat, for example, cannot be conditioned to walk on its two hind legs [3] . Similarly, a raccoon cannot learn to put coins in a piggy bank, as they would have liked to teach him for an advertisement. When the animal held the coins between its paws, in fact, it showed the so-called washing behavior, that is, he rubbed them in a behavior that in nature is useful to remove the exoskeleton of the crustaceans it feeds on [6] . This phenomenon, whereby innate behavior prevails over conditioned behavior, has been called instinctive drift [3, 6] .
Other types of learning
Often behaviors or concepts are imprinted in the memory of animals automatically or passively. These types of learning are distinguished according to the following categories.
Imprinting
First described by Douglas A. Spalding, the phenomenon of imprinting became famous thanks to Konrad Lorenz, who perfected the concept. L ‘ imprinting is a form of irreversible learning (or almost) and innate that takes place during a specific period of life said critical period or sensitive. There are several types [1] .
L ‘ imprinting branch , the most famous, is what allows the individual to distinguish his parents and brothers from other members of the species. It seems that it essentially depends on the recognition of facial features. In one experiment, in fact, chicks in the critical period equally preferred real and stuffed hens, both whole and in pieces (therefore only the head and wings, only the head, etc.), excluding only those that had beak and eyes in the wrong position. On the other hand, hen chicks can also recognize other species of animals as their mother (for example the stone marten , Martes foina ), so this innate predisposition is not very selective [1] .
L ‘ imprinting sex is the one that allows the animal to recognize its own species and be conscious of belonging to it [1] . As in the case of the wild goose ducklings who received the imprinting from Konrad Lorenz: as adults they had a sexual preference towards other human beings [3] .
According to Mark H. Johnson, impriting includes two different mechanisms. The first is the conspec , a predisposition to focus attention on objects similar to one’s conspecifics. For example, our newborns, up to 3 months old, preferentially follow stimuli more similar to faces with their gaze. The mechanism is activated when any face moves in the child’s field of vision. The second is conlearn , the phenomenon by which the individual learns the characteristics of an object . In babies after 3 months of life, for example, the ability to discriminate different faces appears [1] .
L ‘ imprinting has a genetic basis , because not all animals have it in equal measure. For example, if the eggs of Parus major (great tit) are transferred to the nests of Parus caeruleus (blue tit ), they show a much stronger imprinting than the blue tit in the opposite case. The great tit, in fact, will mate only with members of the species with which they grew up, while the blue tit, even if showing this behavior, will also mate among themselves [3] .
Learning-induced releaser
The releaser are, in this case, the stimuli recognized by animals innately that can facilitate the learning of other stimuli . To understand better, let’s see an example. Bees have an instinctive preference for flowers, whose vision stimulates landing and tasting behavioral responses. Consequently, they will also approach what looks like a flower (by color, smell, shape, etc., according to a hierarchy of importance of information). In this way they will more easily learn the characteristics of these elements [1] .
Latent learning
Latent learning occurs when animals learn passively , therefore without being reinforced, and without showing instant changes in their behavior [1] . That is, animals show that they have learned only when what they have learned is useful. An example of latent learning occurs during exploration, as the scientist Edward C. Tolman has shown [7] .
In one experiment he took rats and observed how they behaved inside a maze depending on how hard they were reinforced. One group received reinforcement (food on exit) right from the start, so they almost immediately found their way out. The second group, however, was not reinforced until the twelfth day. For eleven days, therefore, these rats often moved casually, not finding the exit; but at the twelfth they spotted it as well as the rats which had been braced immediately. In this way, the scientist showed that they had learned the way while not receiving any reinforcement [7] .
This ability is also observable, for example, in children who are driven to school. Passively, while not paying much attention to it, they learn the way and show themselves able to follow it when necessary [7] . However, this ability was not believed to be extended to other animals until Tolman’s experiment, which demonstrated that learning is not subordinated to actions [1] .
Insight learning
Sometimes animals learn something suddenly, seemingly out of nowhere, as if they had had an enlightenment . This type of learning was discovered by the psychologist Wolfgang Köhler . He was convinced that animals did not only possess the ability to link their actions with consequences, a belief that spread in the scientific environment following the observations of Skinner and Thorndike. Thus he observed the behavior of anthropomorphic monkeys in conditions different from Skinner boxes: he placed them in large environments containing various elements. He then placed reinforcements in places that could not be reached except with the help of some objects present in the environment in which the monkeys were. For example, he hung bananas at the top[1] .
Initially, the chimpanzees ( Pan troglodytes ) he studied behaved randomly. Then there was a phase of quiescence of variable duration. Finally the animals shook and, without hesitation, solved the problem, for example by stacking crates to be able to reach the reinforcement. This last phase is called insight [1] .
Köhler defined this type of learning as a “cognitive restructuring” of the elements available to the individual. A display of intelligence. Furthermore, once this behavior was shown, the animal never forgot it anymore, putting it into action whenever necessary [1] . Other animals other than primates have also proved capable of this type of learning, such as pigeons .
Social learning
When an individual presents a new behavior it may happen that this dies with him or that it spreads within the population and even that it is also handed down to subsequent generations . In the second case there is social learning , which can take place thanks to different mechanisms: by observing the actions of one’s fellow men ; for cultural transmission; for teaching; or for social facilitation . The latter term indicates the phenomenon by which an individual carries out a certain behavior more frequently in the presence of individuals who carry it out in turn [8] .
Social learning has been observed mainly in primates, such as humans and chimpanzees, but also in some invertebrates, such as bees and ants Myrmica rubra . An example of social learning is that seen in the 1950s in passerines belonging to the Paridi family. In one country in England there were more and more sightings of these animals opening milk bottles.