Learning
Katarzyna Wojczulanis-Jakubas
21 December, 2021
Introduction
(Animal) learning is the process of acquiring new understanding, knowledge, behaviours, skills, (values, attitudes, and preferences, those in brackets hard to establish/examine in animals). In other words, (animal) learning it is the alternation of behaviour as a result of individual experience. When an organism can perceive and change its behaviour, it is said to learn. Although many animal behaviours can be fixed and performed in unvarying routines (e.g. females insect lay their eggs on a particular plant or captured prey, the newly hatched larvae immediately start eating and then follow a standard sequence of developmental stages; etc), there are many examples that animals can learn (e.g. Yuki is playing dead). Thus, the fact that animals can learn is actually not the subject of the discussion. What is interesting is how animals learn, and what is actually animals understanding of the reality; these questions prove exceedingly difficult to answer.
For sure, there are different kinds of learning, and the output may be astonishingly complex. Some learning is immediate, and induced even by a single event; the output is then rather straightforward (e.g. avoiding a fire being once burned by). Much skills and knowledge, however, accumulate from repeated experiences (in the same or different context) and may yield to pretty sophisticated behaviour/s. Then, the learning-induced changes may last lifetime, but may also be “lost” after a period of time (especially if not refreshed, at least episodically).
Learning may occur as a result of imprinting (very basic), non-associative learning – habituation/sensitisation, associate learning - classical or operant conditioning, or as a result of more complex activities (trial-and-error, insight, social learning, teaching).
Imprinting
Imprinting it is phase-sensitive learning (learning occurring at a particular age or a particular life stage; critical period). It is rapid and apparently independent of the consequences of behaviour. In basic form, it occurs in situations in which an animal (or a person) learns the characteristics of some stimulus, which is therefore said to be “imprinted” onto the subject.
Critical period (in developmental psychology/biology) - a maturational stage in the lifespan of an organism during which the nervous system is especially sensitive to certain environmental stimuli. If, for some reason, the organism does not receive the appropriate stimulus during the critical period to learn a given skill or trait, it may be difficult, ultimately less successful, or even impossible, to develop certain associated functions later in life. For example, the first few years of life constitute the time during which first language (determining native-like competence) develops readily and after which (sometime between age 5 and puberty) language acquisition is much more difficult and ultimately less successful (Sigler 2006).
Non-associative learning
Non-associative learning refers to “a relatively permanent change in the strength of response to a single stimulus due to repeated exposure to that stimulus“. This definition exempt the changes caused by sensory adaptation, fatigue or injury. Non-associative learning can be divided into habituation and sensitization.
Habituation
Habituation is an example of non-associative learning in which one or more components of an innate response (e.g., response probability, response duration) to a stimulus diminishes when the stimulus is repeated [habituation is different from extinction, which is an associative process, see below].
An example of habituation can be seen in small song birds—if a strawman is put in the field the birds, like sparrows initially react to it as though it were a real threat (a predator or similarly threatening). Soon the birds react less, showing habituation. The habituation rate depends of the frequency and intensity of the stimuli (it is faster for stimuli that is weak and/or occurs at a high rate). Habituation has been shown in essentially every species of animals. Habituation is in direct opposition to sensitization.
Sensitization
Sensitization is an example of non-associative learning in which the progressive amplification of a response follows repeated administrations of a stimulus This is based on the notion that a defensive reflex to a stimulus such as withdrawal or escape becomes stronger after the exposure to a different harmful or threatening stimulus. For example, a dog that has been treated badly may be “oversensitive” to a simple gesture like rasing a hand.
Another example for habituation and sensitization comes from studies on drug use in laboratory rats. Habituation may lead to necessity to use higher dosage of a drug, to get the same effect. But also usage of some “light” drugs may increase probability of usage of heavy ones (gateaway drug effect) (Herman 2013). Possible causes are biological alterations in the brain due to the earlier substance exposure and similar attitudes of individuals who use different substances across different substances (in humans - common liability to addiction). Some studies suggests study suggest that marijuana is a “gateway” to more dangerous substance use, though not for the majority of people who use substances (Secades-Villa et al. 2015).
Associative learning
Associative learning is the process by which a person or animal learns an association between two stimuli or events. We have here classical and operant conditioning.
Classical conditioning
In classical conditioning, a previously neutral stimulus is repeatedly paired with a reflex-eliciting stimulus until eventually the neutral stimulus elicits a response on its own. The most famous example is Pavlov’s experiment. The researcher fed his dogs with meat, which naturally made the dogs salivate, as salivating is a reflexive response to the food. Meat then is the unconditioned stimulus (US) and the salivation is the unconditioned response (UR, a reflex). Pavlov rang a bell before presenting the food. The first time Pavlov rang the bell, the neutral stimulus, the dogs did not salivate, but once he put the meat in front of the dogs they began to salivate. After numerous pairings of bell and food, the dogs learned that the bell signalled that food was about to come, and began to salivate already when they heard the bell. Once this occurred, the bell became the conditioned stimulus (CS) and the salivation to the bell became the conditioned response (CR). Since then, classical conditioning has been demonstrated in many species (including humans).
Instrumental conditioning
In operant (instrumental) conditioning, a behaviour that is reinforced or punished in the presence of a stimulus becomes more or less likely to occur in the presence of that stimulus. In operant conditioning the behaviour is controlled by external stimuli, and the operant behaviour is considered to be “voluntary” (In contrast, classical conditioning involves involuntary behaviour based on the pairing of stimuli with biologically significant events).
Reinforcement and punishment are the core tools through which operant behaviour is modified. These terms are defined by their effect on behaviour. Either may be positive or negative.
Positive reinforcement and negative reinforcement increase the probability of a behaviour that they follow, while positive punishment and negative punishment reduce the probability of behaviour that they follow.
Another procedure is called “extinction”.
Extinction occurs when a previously reinforced behaviour is no longer reinforced with either positive or negative reinforcement. During extinction the behaviour becomes less probable. Occasional reinforcement can lead to an even longer delay before behaviour extinction due to the learning factor of repeated instances becoming necessary to get reinforcement, when compared with reinforcement being given at each opportunity before extinction.
Operant conditioning was first extensively studied by Edward Thorndike, who observed the behaviour of cats trying to escape from home-made puzzle boxes (see the video). A cat could escape from the box by a simple response such as pulling a cord or pushing a pole, but when first constrained, the cats took a long time to get out. With repeated trials ineffective responses occurred less frequently and successful responses occurred more frequently, so the cats escaped more and more quickly. Thorndike generalized this finding in his the law of effect, which states that behaviours followed by satisfying consequences tend to be repeated and those that produce unpleasant consequences are less likely to be repeated. In short, some consequences strengthen behaviour and some consequences weaken behaviour. By plotting escape time against trial number Thorndike produced the first known animal learning curves through this procedure.
A learning curve is a graphical representation of the relationship between how proficient an individual is at a task and the amount of experience it has. Proficiency (measured on the vertical axis) usually increases with increased experience (the horizontal axis), that is to say, the more the individual performs a task, the better it gets at it.
Trials and errors
Trials and error is a fundamental method of problem-solving. It is characterized by repeated, varied attempts which are continued until success, or until the practicer stops trying. This approach can be seen as one of the two basic approaches to problem-solving, contrasted with an approach using insight and theory of mind. Biological evolution can be considered as a form of trial and error. Random mutations and sexual genetic variations can be viewed as trials and poor reproductive fitness, or lack of improved fitness, as the error. Thus after a long time ‘knowledge’ of well-adapted genomes accumulates simply by virtue of them being able to reproduce.
Latent learning
Latent learning is a form of learning in which there is apparently no immediate reward (or punishment) for the animal, and what is learnt remains `latent’. The prime example is an animal exploring its surroundings. Learning about the geography of its home area may bring an animal no immediate benefits, but can prove vital in the future when fleeing a predator or searching for food. For example, rats will explore a maze without apparent reward or punishment and will remember the information they obtain; later, when rewards are presented, those rats will perform better at finding their way through the maze than rats with no exploratory experience. Edward Tolman (1948) coined the term cognitive map, which is an internal representation (or image) of external environmental feature or landmark. He thought that individuals acquire large numbers of cues (i.e. signals) from the environment and could use these to build a mental image of an environment (i.e. a cognitive map).
A cognitive map (sometimes called, but should not be confused with, a mental map or mental model) is a type of mental representation which serves an individual to acquire, code, store, recall, and decode information about the relative locations and attributes of phenomena in their everyday or metaphorical spatial environment. The concept was introduced by Edward Tolman in 1948. The concept was used to explain the behavior of rats that appeared to learn the spatial layout of a maze, and subsequently the concept was applied to other animals, including humans.
Insight
Under Morgan’s Canon, many animal behaviours can and should be explained in the simplest possible way. Often, where behaviour seems to imply higher mental processes (insight), it might be explained by “simple” trial-and-error learning (principle of parsimony/Occam’s razor) or latent learning.
Conwy Lloyd Morgan (1852 – 1936) - a British ethologist and psychologist, remembered (among others) for Mogran’s Canon, a principle that played a major role in behaviourism, insisting that higher mental faculties should only be considered as explanation of an animal behaviour if lower faculties could not explain it.
Occam’s razor - entities should not be multiplied beyond necessity”, sometimes inaccurately but memorably paraphrased as “the simplest solution is the usually the best one”. In origin, the rle is that when one is presented with competing hypotheses about the same prediction, should select the solution with the fewest assumptions.
Insight is manifested by sudden solution of a new and difficult task (so-called Aha! moments) can lead to remembering behaviour that is effective in a given situation and with a given drive. Because insight itself is a reorganization of previous experience, learning based on it is not one-sample learning. Insight gives new information to experience gained previously using other types of learning - latent or trial and error. Insight learning (perceiving cause and effect relationships) is the basis for animals’ ability to avoid obstacles. Insight is viewed high mental ability, in the past being considered solely a human trait. Constantly growing number of studies indicates it is not really the case.
A good example of the insight are first studies on chimps by Koehler (see the original video, and more modern version). There are much more of those, and we will soon learn more about that in the cognitive abilities chapter.
Stages of learning by insight:
preparation - exposure to the problem, the state just before frustration.
“incubation” - distancing oneself from the problem, by temporarily “dropping it”.
insight - eureka, enlightenment.
verification - testing the solution and accumulating experience, buffering in memory (if the solution works).
The eureka effect (also known as the Aha! moment, eureka moment, ephiphany, insight) refers to the common human experience of suddenly understanding a previously incomprehensible problem or concept. Some research describes the Aha! effect as a memory advantage, but conflicting results exist as to where exactly it occurs in the brain, and it is difficult to predict under what circumstances one can predict an Aha! moment. A recent theoretical account of the Aha! moment started with four defining attributes of this experience: 1) appears suddenly; 2) the solution to a problem can be processed smoothly, or fluently; 3) elicits positive affect; 4) a person experiencing the Aha! moment is convinced that a solution is true. These four attributes are not separate but can be combined because the experience of processing fluency, especially when it occurs surprisingly (for example, because it is sudden), elicits both positive affect and judged truth.
Teaching
The same mchanisms involved in inadvertent social learning are also employed during teaching. The distinction is drawn based upon the role of the demonstrator. From the perspective of the pupil, teaching would be identical to its inadvertent social learning equivalent, but in teaching, a tutor actively demonstrates a behavior pattern or draws attention to a location with the specific function of transmitting information to the pupil. An individual must meet three criteria to qualify as a “teacher”: it modifies its behavior only in the presence of a naive observer, it incurs some cost (or at least, no benefit) to itself in doing so, and the naive observer acquires knowledge or skill more rapidly or efficiently than it might otherwise.
Until recently it was widely believed that animals did not actively facilitate learning in others. Teaching was regarded as a uniquely human faculty. However, recent studies suggest that teaching might be more common in animals than previously thought. Teaching is present in bees, ants, babblers, meerkats and other carnivores but is absent in chimpanzees, a bizarre taxonomic distribution, that makes sense if teaching is treated as a form of altruism (Hoppitt et al. 2008).
Insects may be good example here. They have been observed demonstrating various forms of teaching in order to obtain food. Some ant species will guide each other to food sources through a process called tandem running, in which an ant will guide a companion ant to a source of food.
Even better example here provide meerkats, where older individuals teach youngs how to handle the food, like scorpiones, etc (Thornton and McAuliffe 2006).
Literature cited
Aplin LM, Sheldon BC, Morand-Ferron J (2013) Milk bottles revisited: Social learning and individual variation in the blue tit, Cyanistes caeruleus. Anim Behav 85:1225–1232. doi: 10.1016/j.anbehav.2013.03.009
Carouso-Peck S, Goldstein MH (2019) Female social feedback reveals non-imitative mechanisms of vocal learning in zebra finches. Curr Biol 29:631-636.e3. doi: 10.1016/j.cub.2018.12.026 Herman JP (2013) Neural control of chronic stress adaptation. Front Behav Neurosci. doi: 10.3389/fnbeh.2013.00061
Hoppitt WJE, Brown GR, Kendal R, et al (2008) Lessons from animal teaching. Trends Ecol Evol 23:486–493. doi: 10.1016/j.tree.2008.05.008 Jelbert SA, Taylor AH, Cheke LG, et al (2014) Using the aesop’s fable paradigm to investigate causal understanding of water displacement by new caledonian crows. PLoS One 9:. doi: 10.1371/journal.pone.0092895
Secades-Villa R, Garcia-Rodríguez O, Jin CJ, et al (2015) Probability and predictors of the cannabis gateway effect: A national study. Int J Drug Policy 26:135–142. doi: 10.1016/j.drugpo.2014.07.011
Siegler R (2006) How Children Develop, Exploring Child Develop Student Media Tool Kit & Scientific American Reader to Accompany How Children Develop. New York: Worth Publishers
Thornton A, McAuliffe K (2006) Teaching in wild meerkats. Science (80- ) 313:227–229. doi: 10.1126/science.1128727
Tolman EC (1948) Cognitive maps in rats and men. Psychol Rev 55:189–208. doi: 10.1037/h0061626
Whiten A, van de Waal E (2018) The pervasive role of social learning in primate lifetime development. Behav Ecol Sociobiol 72:. doi: 10.1007/s00265-018-2489-3 Zhang Y, Yu C, Chen L, Li Z (2021) Performance of Azure-winged magpies in Aesop’s fable paradigm. Sci Rep 11:1–11. doi: 10.1038/s41598-020-80452-5
Social learning
Social learning (imitation)- the performance by an individual of new and not genetically programmed actions, similar to those observed currently or previously in other individuals. In other words, imitation is the memorising of someone else’s behaviour, and this usually takes place in the presence of a corresponding drive. Further consolidation of the imitated skills in the course of one’s own actions is based on trial and error or motor learning. This type of learning is also optional. Well developed in mammals, but also in birds (singing, foraging) and other (often social but not necessarily) animals.
Some behaviours are learnt via social learning in few phases, see primates for example (Whiten and van de Waal 2018), but social learning is likely to be a life-time process.
Avian examples, foraging - great and blue tits are famous for the ‘milk bottle’ innovation, which emerged at numerous sites across Britain, and spread owing to social learning, in the early 20th century. This is also a good example to note that there is a great variation among individuals in learning performance, which may be related to age, sex, and/or behavioural profile of individuals Alpin et al. 2013. More on that in the chapter on inter-individual variation.
In some cases the social learing can be further guided/enhanced. A good example of it are zebra finches. Overall, the speceis is the most common model of human speech development but, unlike humans, is thought to learn only via imitation/social learing. Recent study of Carouso-Peck and Goldstein (2019) shows that song learning is affected by non-vocal, visual feedback from females (see the video of female behaviour). Young males given feedback contingent on their immature songs developed higher-quality songs than yoked controls.
Social learning is a basic process behind evolution of animal culture.
The concept of animal culture involves the current theory of cultural learning in non-human animals, through socially transmitted behaviors. The question as to the existence of culture in non-human societies has been a contentious subject for decades, largely due to the lack of a concise definition for the word ‘culture’. However, many leading scientists agree on seeing culture as a process, rather than an end product. This process, most agree, involves the social transmittance of novel behaviour, both among peers and between generations. Such behaviour can be shared by a group of animals, though not necessarily between separate groups of the same species.