Animal Learning and Cognition Revision Notes PDF

# Animal learning and cognition revision notes ## Historical background * Darwin's theory- evolutionary continuity between animals and humans * Natural selection * There is variability among individuals in a population * Variability is hereditary * Organisms produce more offspring than...

# Animal learning and cognition revision notes ## Historical background * Darwin's theory- evolutionary continuity between animals and humans * Natural selection * There is variability among individuals in a population * Variability is hereditary * Organisms produce more offspring than the environment can sustain, leading to competition for limited resources * Survival of the fittest ### Romanes (1881) * Anthropomorphism- the tendency to view animals as people or to have the same capabilities, motivations and desires * Animals are rational, empathetic, reasoning creatures * Simple progression in the intellectual capabilities of animals- brought about by natural selection * Insects, Birds, Apes, Humans ### Llyod Morgan (1890s) * Argued against anthropomorphism * Dog opening a gate was due to trial and error, not higher intelligence ### Lloyd Morgan's (1894) Canon In no case may we interpret an action as the outcome of the exercise of a higher psychical faculty, if it can be interpreted as the outcome of one which stands lower in the psychological scale ### Thorndike (1874-1949) * Studies animal learning * Puzzle boxes with cats * Wanted to deliver the coup de grace to the despised notion that animal's reason * Thorndike's law of effect ## Stimulus response association >- The outcome (O) only strengthens or weakens the connection between S and R, and the outcome itself is not represented in animals on making the instrumental response - If a response leads to a satisfying outcome, it will be strengthened and will be more likely to occur in the future ## Watson (1878-1958) * Importance of learning * Studied unconscious processes * Behaviourism- humans are blank slates- explains behaviour in terms of the environment * Argued against introspection ## Titchener * Structuralism and introspection (shared ideas with Wundt) ## Skinner and Pavlov ### Skinner (1904-1990) * Radical behaviourism (applications of findings to humans) * Studied animals that could be kept in a lab * Operant/ instrumental conditioning * Response shaping, schedules of reinforcement * Stimulus response association ### Skinner box- automated apparatus- conditioning chamber #### Interval schedules * Presenting a reinforcer after a certain amount of time ###### Fixed schedule * The first response after a fixed amount of time (interval) or #### Ratio schedules * Presenting a reinforcer after a certain number of responses ###### Variable schedule * The time of responses varies (interval) The number of responses reinforced responses (ratio) varies (ratio) ## Pavlovian/ classical conditioning Pavlov plotted the number of drops of saliva (CR) produced over successive CS-US pairings on a graph. During Acquisition the number of drops increased over trials, up to a stable maximum number (Asymptote). When the CS was presented alone over a number of trials, conditioned responding dropped off to almost nothing. This is known as **Extinction**. However, after a 24 hour rest the re-presentation of the CS alone produced a CR. This is known as **Spontaneous Recovery** and is evidence that in Extinction the relationship between the CS and the US is not simply forgotten, because Spontaneous Recovery would not otherwise occur. There was lower responding as the frequency of the tone deviated more from the CS Plotted on a graph this is known as a **Generalisation gradient** and is the result of **Stimulus generalisation**. When conditioned responding is lower due to a slight difference between the trained CR and the presented CR (such as a change from 1200 to 1000Hz) this is known as **Generalisation decrement**. Acquisition, Extinction, Spontaneous Recovery and Generalisation decrement have all been shown in human studies such as the Little Albert study by Watson, and the sexual fetishism study by Rachman ## Animal intelligence Nakajima et al. (2002) asked students to rank the intelligence of various animals on a scale from 0 to 100. Linear progression from invertebrates->fish->chickens->sheep-> cows->horses-> cats> dogs-> great apes ### Aristotle's great chain of being * Sequence of complexity- sponges ---> humans ### Brain size * Species with a bigger brain possess a greater potential for intelligence * Jerison- ratio of brain weight to body weight (cephalisation index) A larger cephalisation index indicates a big brain for the size of the animal's body, and a low number indicates the reverse relationship. The principle of proper mass- the more important a function is, the more of the brain is devoted to it. Problem- possible for other factors such as boldness or neophobia to contribute to innovation (factors which are not involved in intelligence) ### Foraging intelligence hypothesis- * specific brain regions might correlate with some behaviours, such as the size of the hippocampus and food storing in birds ### Social intelligence hypothesis- * social complexity is the primary factor that drove the evolution of large brains and advanced cognitive abilities in animals Anaxogoras- all animals were equally intelligent, but some were more expressive than others. Macphail- results could be explained by differences in factors like motivation, perception or motor skills rather than differences in intelligence. ## Learning * permanent change in behaviour resulting from experience rather than the acquisition of knowledge, which is a mentalistic process that cannot be observed in animals ## Problems with speed: * difficult to determine the motivation of each animal to carry out the task * Skard- compared rats and humans in their ability to navigate through a maze * Angermeier- trained various animals to make a response to gain a food reward- fish were the quickest, followed by chickens, rats and rabbits * Opposite to results on the cephalisation index ## Contextual variables * Difficult to equate the **PERCEPTUAL demands of the test** * Difficult to equate the **MOTIVATIONAL demands of the test** ## Bitterman- systematic variation Conducted a series of experiments where the perceptual and motivational demands of the task were systematically varied. ## Garica and Koelling (1966) * Trained thirsty rats to drink from a water dispenser Experiment Setup: Rats were exposed to a bright light and noise while drinking water. This combination was intended to act as a conditioned stimulus. Conditioning: The rats were then subjected to either radiation or a toxin (like lithium chloride) paired with the bright light and noise. Results: The rats developed an aversion to the water when it was paired with radiation or a toxin, but not when paired with the bright light and noise alone. Conclusion: This suggested that biological predispositions influence learning, as the rats more readily associated taste (a gustatory stimulus) with illness rather than a visual or auditory stimulus * Speed of learning might direct attention away from other important intellectual capacities ## How do we measure animal intelligence? * Cognition and reasoning- thinking, knowing and understanding * Ability to process information in an adaptive way * Can be modelled in terms of computations ## Animal memory ### Vander Wall (1982) Living in high-altitude pine forests, the Clark's nutcracker faces seasonal food scarcity. It adapts by storing about 30,000 seeds across 4,000 caches and uses memory to find them later- Similar memory abilities have been demonstrated in pigeons. ### Vaughan & Greene (1984) pigeons learned to peck a key for some photos but not others. The birds successfully learned to discriminate between 320 different photographs, demonstrating impressive memory capacity. ### Time periodic Animals can remember temporal information through periodic timing - responding at specific times. For example, mice show daily activity cycles even under constant conditions, while cockroaches follow a 23.5-hour cycle. These circadian rhythms are influenced by environmental cues like sunlight and temperature, though there are limits to how much these cues affect behaviour. ### Bolles & Lorge (1962) Rats raised with 19- or 29-hour cycles couldn't learn to anticipate feeding times, suggesting the 24-hour clock is innate. ### Church & Deluty (1977) Rats were trained with a two-lever system. Pressing the left lever after a 4-second tone interval earned food, while the right lever earned food after 16 seconds- demonstrates rats' ability to remember time intervals. ### Hans In the late 1800s, Clever Hans, a horse famous for solving math problems, would tap his hoof to count and stop at the correct answer. Pfungst (1908) discovered Hans wasn't actually counting - he was reading his questioner's body language, stopping when they relaxed at the right number. ### Meck & Church (1983) rats were trained with 4 versus 16 tones linked to left and right lever rewards. When tone presentation time was equalized, rats still chose correctly, demonstrating they learned to count rather than time the stimuli. ### Brannon & Terrace (2000) Showed that monkeys were able to remember the numerosity represented by particular stimuli, by training them to touch squares on a touchscreen with different numbers of dots on them. The monkeys had to touch the squares in numerical order, as shown by the number of dots, to gain a food reward. The experimenters ruled out the possibility that the monkeys were solving the problem by remembering differences in the brightness of the stimuli, or by simple template-matching (like taking a visual snapshot), by conducting a number of test trials that controlled for these potential confounding factors. ### Rugani, Regotin and Vallortigara (2007) Showed that five-day-old chicks are able to remember the numerical serial position of a baited food well, in a row of nine other unbaited wells. Different chicks were trained to find food in the third, fourth, or sixth food well in the row. The experimenters ruled out the possibility that the birds had simply learned the spatial location of the correct food well by rotating the row through 90 degrees and testing the birds from a new starting position. ### Parrot- Pepperberg (1994) performed a variety of tasks- subitizing- can only name the number of items in tasks, up to a max of 7 or 8 items. ### Herrnstein, Loveland & Cable (1976) The results of this study led Herrnstein et al. to propose that pigeons had the ability to acquire concepts, or categorise information. ### Cerella (1979) Pigeons were required to categorise silhouettes of oak leaves from non-oak leaves, lends support to this idea. ### Explanations of Categorisation 1. Innate categories- characteristics or abilities that are present in an organism from birth 2. Exemplar learning- people categorize new objects by comparing them to examples they already know. 3. Feature learning- extraction of distinctive features from data for various tasks such as classification, anomaly detection, and clustering 4. Exemplar learning with stimulus generalisation- responding to stimuli that are similar to the original conditioned stimulus ## Memory and intro to associative learning ### Habituation A reduction in responsiveness to a stimulus as a result of its repeated presentation ### Whitlow (1975) (rabbits) plethysmograph Presented a loud tone to rabbits in a chamber and measured the change of blood flow in their ears in response to each presentation of the tone. He found that on the first presentation (S1) there was a strong response, but on the second presentation (S2), after a 60 second delay, there was a weaker response. If the interval was increased from 60 to 150 seconds the response to S2 was just as high as the response after S1. ### Effector Fatigue If the physiological system that is responsible for overt behaviour is fatigued then it will not be able to respond to any stimulus within the period of fatigue. ### Receptor Fatigue Cells responsible for the reception of S1 might be fatigued after the offset of S1 and would therefore be less sensitive at the presentation of S2. ### Radial Maze- Olton (1978) Animal is allowed to visit four out of eight arms on the maze. At the end of each visited arm is a food reward. In between each visit to an arm the animal returns to a central hub. Rats are readily able to complete this task with 8- or 12-arm mazes, but find the task difficult with 17 arms. ### Retroactive interference ### Beatty & Shavalia (1980) Trained rats to complete a normal radial maze task, with a one-hour delay after the fourth arm had been entered. In the interval the rats were required to complete a second radial maze task in a different room. This second task appeared to have no effect on the rats' memories for the first task. ### Roberts (1981) (3 radial arm maze tasks- retroactive interference occurs) ## Mechanisms of LTM * Consolidation * Retention * Retrieval ### Hebb (1949) Consolidation and rehearsal is needed to complete these connections→ memory depends on the virtually permanent formation of circuits of interconnected neurons. ### Duncan (1949) ECS impaired the memory and rehearsal task. ### Holland (1990) CS is able to activate some of the perceptual mechanisms that are normally activated by the US. * USs have Specific and Affective qualities * Specific: things that are unique about the stimulus e.g the flavour e.g wintergreen vs peppermint * Affective: common similarities between the US and other USS * Affective qualities can be Appetitive or Aversive * Appetitive: Food, water, and the opportunity to mate are all things that might be said to lead to a state of satisfaction in the animal * Aversive: Mild electric shocks, illness and loud noises are all things that might be said to be unpleasant stimuli, that the animal might try to avoid if possible- stimulus to avoid a shock ### Autoshaping starts to peck at the light (CS) behaves as it is the US (food or water) ### Wagner's theory Learning can be explained in terms of associations between conditioned and unconditioned stimuli ### Mackintosh's theory Animals will pay attention to and learn readily about stimuli that are good predictors of significant events, such as food or shock. In other words, stimuli with high associative strength are likely to receive more attention than those with low associative strength. To be attended to fully, the CS must not only be a good predictor of the US, it must be a better predictor than all the other stimuli that are present on a trial. ### The Pearce-Hall theory Associative learning- a stimulus gains more attention and is more readily learned when it is paired with a surprising or unexpected outcome. ### LePelley's theory We tend to focus on cues that tell us what to expect next based on past experiences. Blocking will be a consequence of the interaction of two factors: * Loss of effectiveness by the US, because it is no longer surprising * Loss of attention and associability, by the added stimulus because it is a relatively poor predictor of the US. ## Associative learning ### Biological * Metabolic changes * Synaptic changes ### Psychological * Conditioning affects homeostasis mechanisms ## Contuiguity- events that occur close together in time or space will become readily associated.