Psychological Foundations of Mental Health Week 1 Introduction to Cognitive Psychology PDF

Module: Psychological Foundations of Mental Health Week 1 Introduction to cognitive psychology Topic 1 Foundations of cognitive psychology from Plato to Pavlov - Part 1 of 3 Professor Richard Brown Department of Psychology, King’s College London Lecture transcript Slide 3 Let’s start with some basic definitions. As in all disciplines, psychology has its own jargon-- technical words developed and devised specifically for the communication of ideas and information to other psychologists. However, we also use words and terms that are in everyday use but use them in a more specific way. But this can sometimes cause confusion, so it’s important always to be aware of definitions. The word “cognition” has both an everyday and a specific definition. In its everyday use, a cognition is another word for an individual thought. In psychology, however, it has a more specific and more detailed definition and usage. It is a term applied to all forms of mental processes-- conscious and unconscious, deliberate and automatic. The term covers processes involved in a range of areas or domains of function including perception and memory, the understanding of language, how we identify the objects in our world, how we form and use concepts, interpret events, ascribe meaning, make judgments and decisions, solve problems, plan, and so on-- essentially, everything that allows us to function within the world. This list and more forms what psychologists place in the category of cognition. Slide 4 It follows, therefore, that cognitive psychology is the branch of the subject devoted to the scientific study of these mental and cognitive processes. Cognitive psychology seeks to identify and understand the internal representations and structures that underlie our conscious and unconscious cognitions. We will come back to these terms and what they mean in more detail later in the week. Cognitive psychology is based on building theoretical descriptions or models of cognitive structures and processes. Because we cannot observe them directly, the structures and processes can be considered hypothetical constructs. Their existence in nature is inferred from a combination of testable theory and experimental study. This is the definition of “cognitive psychology” that we will use in the programme. Cognitive psychology is one branch of psychology-- although arguably, the dominant school of the later 20th century and 21st century. Transcripts by 3Playmedia Week © King’s College London 2018 1. Slide 6 Before moving on, it is important to note that cognitive psychology is just one of a group of related disciplines collectively called “cognitive science”. When we talk about cognitive psychology, we are mainly referring to experimental cognitive psychology. However, its subjects of study-- such as attention, memory, decision-making, and so on-- have, since the mid-20th century, also been the focus of other emerging areas of science. These include the study of brain mechanisms and functional cognitive neuroscience, computers and artificial intelligence, and psycholinguistics. These different areas, once largely separate, have grown closer and closer-- building on each other’s theories, methods, and findings to form evermore integrated models of the mind and how it works. This integration is often called the “cognitive revolution.” It started in the middle of the 20th century but continues to this day. In this programme, across the various modules, you will be learning mainly about two of these areas-- experimental cognitive psychology and cognitive neuroscience and their relevance to mental health. Transcripts by 3Playmedia Week © King’s College London 2018 2. Module: Psychological Foundations of Mental Health Week 1 Introduction to cognitive psychology Topic 1 Foundations of cognitive psychology from Plato to Pavlov - Part 2 of 3 Professor Richard Brown Department of Psychology, King’s College London Lecture transcript Slide 2 How did we get to our current state of knowledge and understanding? This timeline shows just a few of the key periods and the figures that have contributed to making cognitive science and cognitive psychology what it is today. In this first topic, we are going to travel along this timeline, covering the first 2,000 or so years rather quickly, before slowing down as we hit the 19th and 20th centuries, when psychology as we recognise it began. We will trace some of the key influences in the emergence of psychology as a scientific discipline and the splitting of cognitivism and behaviourism, before we see a recognisable cognitive psychology emerge in the mid-20th century. A word of warning. This early history is dominated by men and, for much of the time, men with impressive beards. Slide 4 We start over 2,000 years ago. The philosophical traditions and schools of thought that have shaped modern cognitive psychology have their origins in ancient Greek philosophy, from figures such as Plato and Aristotle. The concept of psychology did not exist in the same way as we consider it today. However, from its foundation, philosophy was fundamentally concerned with the working of what we would call the mind and particularly the human mind. The modern word “psychology” derives from the Greek word “psyche,” which refers to the soul as distinct from the physical body. However, in the thinking at that time, the soul encompassed a wide range of concepts, including what we would call the mind. Two broad schools of thinking emerged, called rationalism and empiricism, that still influence and echo through modern psychological theory today and are seen as complementary, rather than opposing. The rationalist schools of philosophy, such as that proposed by Plato, broadly held that we can explore the mind and other abstract ideas and constructs through a process of thinking itself, by examining personal experience and, through that, mental processes. These include intuition, which can be considered a form of rational insight arising from exploration Transcripts by 3Playmedia Week © King’s College London 2018 1. of an idea, and the process of deduction, where we use logical processes of reason to draw novel, general conclusions from existing knowledge and experience. Rationalist thinking also often presumes that much knowledge, including mathematical concepts such as quantity and logic, is innate or acquired through development, without the need for it to be learned. This is a version of the nature view in the nature versus nurture dichotomy. This view also led to a broader generalisation that we all have an invariable core, our individual human nature, that cannot be altered or manipulated. It is this that makes each of us unique, whether good or bad. The empiricists, such as Aristotle, a pupil of Plato, adopted a very different approach. First, they believed that the fundamental concepts of the mind and all knowledge comes from sensation and experience, rather than sitting there, waiting to emerge. Also, we are shaped by what happens to us, the nurture viewpoint. Later empiricists often had a sceptical distrust or more fundamental objection to the unobservable, preferring evidence of our senses and what emerged from them. There was also a profound preference for simpler explanations, ones that were sufficient to account for what could be experienced and observed. Such simplicity was contrasted with explanations based on the internal and the unobservable. These tended to be more complex, arising from the rationalist exercise of intuition and reason alone. Finally, an implication of the empiricist view was that human beings can be controlled and manipulated exceptionally easily. We are nothing other than what we experience and so can be influenced to do whatever we’re taught, again, for good or bad. Such empiricist and rationalist approaches, in many forms, have coexisted for thousands of years, sometimes peacefully and sometimes as bitter battles between rival factions. Today, rationalist and empiricist principles are not seen as contradictory or conflicting. Rather, they can be applied to different methods of inquiry and subject areas. The modern scientific method involves the interplay between reasoning and abstract theorising, from which generalisations arise by deductive processes, and induction, where we use empirical methods, such as observation and experiments, to test the validity of those theories and so refine them. Similarly, the nature versus nurture debate that has raged for thousands of years is now seen as scientifically sterile and unhelpful, as well as fundamentally inaccurate as a dichotomy. Slide 5 Let’s fast forward almost two and a half thousand years to the so-called European Age of Enlightenment and the start of the Scientific Revolution. Kant was a German, 19th century philosopher and one of the most influential figures in modern philosophy and arguably one of the founding influences on modern psychology. However, he could just as easily have killed it off before it started. He worked in an age of enormous growth in scientific understanding of the natural and physical world, in chemistry, in physics, and biology, through the use of precise measurement and experimental method to create mathematically definable descriptions and laws. In many of his works, especially in a Critique of Pure Reason, published in 1781, he addressed the issue of whether psychology, the study of the mind, could ever be an empirical science in the same way as the physical sciences, such as chemistry and physics, which he felt were governed by the immutable laws of mathematics, or whether it must remain within the realm of philosophy or natural science. He concluded, albeit with some self-argument and contradictions, that it was not and could not be an empirical science, that the mind and its functions were not amenable to direct study. He based this conclusion on a number of arguments. First, he believed that we can only rationally study our own thoughts and internal processes and Transcripts by 3Playmedia Week © King’s College London 2018 2. not those of others-- in other words, that the method of psychology must be introspection, the process of examining our own thoughts and feelings. However, he also reasoned that one person’s introspection may be very different from another’s and, therefore, a poor method to reveal any underlying laws governing the mind. He also reasoned that introspection artificially forces us to separate things that may not be separable, that such so-called reductionism may lead to false conclusions when it comes to studying the mind. Finally, he argued that the process of introspection alters what we are attempting to observe and understand. So rather than challenging his first premise and concluding that the method, introspection, was fundamentally flawed and looking for another empirical method, Kant concluded that the scientific study of the mind was doomed. Goodbye, psychology. Why then do we consider Kant an important figure in the history of cognitive psychology? Well, first, he defined the mind as a set of separate abilities or functions, but which work together as a whole to produce our experience and onto the level of knowledge and understanding. Essentially, he described what we would now call a cognitive architecture, linking elements of perception, cognitive transformation, and knowledge. It’s described in one of his many famous quotes, as shown here. Second, although rejecting the possibility of direct study and measurement, he proposed that, even if we cannot observe the mind, we can infer the conditions that must be present in the mind to explain our conscious experience, what he called a transcendental method. He viewed this as a philosophical rather than a scientific method. However, in suggesting that there are ways to infer what we cannot directly measure, he established the basic tenet of modern cognitive psychology. Today, fortunately, we have a wide range of empirical methods to test the validity of those inferences. Fortunately, Kant did not kill off psychology. And it continued to develop over the early half of the 19th century in Europe and later in the USA. Over that time, various schools of psychology developed, driven by different underlying philosophies and schools of thought. Although none of these are dominant influences today, they have all contributed, in part, to what we recognise today as cognitive psychology. Slide 6 Wilhelm Wundt is often credited as being the father of experimental psychology. He set up one of the first ever experimental psychology laboratories in Leipzig, Germany, in 1879. Like Kant, Wundt believed that introspection was the most direct way to study the conscious mind, the only sort that he felt important. But he sought to make it an empirical tool fit for purpose, with his techniques of experimental introspection. In his experimental method, he concerns himself primarily with introspection based on the sensations and the percepts, feelings, or thoughts that arose directly from them, the latter particularly in the form of visual images. After a warning, he presented stimuli, such as objects, sounds, words, and pictures for a brief but precisely controlled period of time, using electronically controlled apparatus in a dark and quiet room. The stimuli themselves were systematically varied to measure the effect this had on the resultant internal cognitive event. Those doing the introspection were highly trained to attend to and describe their experiences without interpretation, with the goal being to understand how the initial exposure led first to automatic and what Wundt called passive associations-- in other words, how they had registered or attended to the presence of a stimulus, such as a sound, and its basic properties. There was then the process of how that led to the conscious thought, when the precise nature of the stimulus was registered. Wundt’s particular focus was on the formation of a mental image to the stimulus, a process Wundt called apperception This was seen to result from an active, Transcripts by 3Playmedia Week © King’s College London 2018 3. controlled, and voluntary process, and represented the higher function of the mind. His school of thought is sometimes referred to as voluntarism. This separation of automatic and controlled cognitive processes is still important today, although, unlike Wundt, we no longer see the controlled ones as inherently superior. Slide 7 Wundt was interested in looking at conscious experience as a whole, rather than understanding the individual nature of the initial sensation or the resultant image. In a typical experiment, a subject might be shown a word, such as “apple”, and might immediately report an associated property, such as crisp or sweet or red. This will be reported as preceding conscious awareness of the whole object, the apple itself. From such results, Wundt reasoned that our thoughts had structure, that when the mind was presented with an apple, it received a set of properties that defined its structure before being combined to form the thought or mental image, a principle called structuralism. Wundt attempted to measure these processes further, through the use of electromechanical devices to record reaction times from the presentation of the stimulus to the introspective response. In this way, he sought to separate out the time course for his core processes of attention and apperception an approach he called mental chronometry. His trained researchers would react as quickly as they could by pressing a button, either when they had registered the stimulus or when they had formed the image. He used the so-called subtractive procedure, the time difference between the two different introspective products, to estimate the extra time needed for apperception to take place, he estimated to be about 1/10 of a second or 100 milliseconds. Significantly, Wundt did not attempt to study what we would now recognise as core aspects of cognition, such as learning and memory, language, or reasoning. This was not because he denied that they existed or felt that they were unimportant, more that they were less accessible to his experimental methods and to introspection. Thus, Wundt limited his methods and theories only to what could be studied. Slide 9 Edward B Titchener was born in the UK and studied with Wundt in Leipzig, before moving to Cornell University in the US. Like Wundt, he used the experimental approach of introspection. But unlike Wundt, who was interested in the final, whole, conscious experience, Titchener was interested in using experimental introspection to reveal and understand the elemental parts of conscious experience, by breaking them down, an approach that his teacher, Wundt, strongly and vocally disagreed with. Titchener’s approach to psychology was fundamentally structuralist, taking the basic principle of Wundt to extremes and extending it greatly. At the same time, he was not greatly interested in holistic processes involved in properties of the mind, such as the will. Titchener’s experimental methods were described in his two volume book, Experimental Psychology, a Manual of Laboratory Practice, published in 1902. Even more than Wundt, he sought to standardise the experimental method of introspection to improve its accuracy and reproducibility. Titchener’s psychology and the methods used was very much focused on the immediate and mental experience, the sensation, rather than what use was made of that sensation or the function it served. In other words, how the brain answers the question, what is it? Rather than, what is it for? He believed that cognitions arose from sensations. And so we need a science of sensation as the Transcripts by 3Playmedia Week © King’s College London 2018 4. essential foundation to understand the mind. He proposed that every sensation, whether sound or smell, was made up of four independent properties-- its intensity, quality, duration, and spatial extent. For example, a presented image could be bright, its intensity; pale red, its quality; brief, its duration; and small, its extent. These properties could then each be sub-classified further. He applied his model of reductionist classification to a fantastic degree and, in an earlier book, suggested a list of over 40,000 elemental qualities that would be worthy of study. Here’s a quote that gives a flavour of Titchener’s structuralist thinking. Slide 11 For the structuralist, presented with an apple, the aim was to identify the key sensory features that defined apple-ness and so permitted the identification of the apple as an object. That same set of features would also permit discrimination between one fruit and another or between an apple and a ball. Whether or not an apple is simply defined as the sum of its sensory parts is debatable. But the idea of what we would now call feature analysis, as a means of object identification, remains in more contemporary cognitive theories. However, feature analysis is typically seen only as the first step in a range of ever more cognitive and independent stages of analysis and decision making. There is more to making a decision to pick up an apple and eat it than breaking down its features into their component parts and coming up with “apple.” However, the structuralist approach as a primary model of psychology was fundamentally limited by its focus on sensation. Like Wundt and voluntarism, Titchener’s psychology was defined by what we could see and measure. Although cognitive in some respects, it neglected whole areas of what we now consider to be contained within the realm of cognitive science-- memory, perception, decision making, and so on, let alone areas such as consciousness and the will. Slide 12 Around the same time as Wundt, Titchener, and others were working in Europe, using introspection and experimental methods to study the structure of the mind, another important influence on modern psychology came out of the USA from the writings of William James, brother of the novelist Henry James. William James brought together a disparate body of work in his seminal, two volume work, Principles of Psychology, published in 1890, taking him 12 years to write. He authored the first ever university course in psychology in the US at Harvard University. Unlike Titchener and Wundt, he was not a great experimentalist and complained about the tedium of lab work. Freed from the constraints of the experimental methods of the time, James’s writings could range wider and include a consideration of psychological functions such as consciousness, instinct, and emotion. Like others, he used introspection to inform his theories. However, in contrast to the experimentally driven structural approaches, James adopted a functional approach. Informed by Darwin’s evolutionary theory, he attempted to explain how the mind adapts to meet the needs of the organism. As a proponent of the functionalist school of psychology at the time, he was interested in mental processes, rather than mental structures-- for example, what was the purpose of consciousness or the will, rather than trying to dissect and define its constituent parts. To use a crude analogy, a structuralist who wants to understand how a car worked would stop it, break it down into its parts, and attempt to work backwards from that knowledge. A functionalist would drive the car and you use that experience to understand how its constituent parts worked together and allowed it to move, start, stop, and turn. Again, James’s methods have long since been replaced by modern experimental approaches. Transcripts by 3Playmedia Week © King’s College London 2018 5. However, functionalism remains a core part of modern psychology, with its focus on cognitive processes and how would they serve adaptive behaviour. However, it has merged with other aspects of the structuralist tradition to form an overarching framework to make up cognitive psychology. Slide 14 At the same time as the introspective approach held dominance within psychology, others were applying experimental methods to understanding more directly important properties of the brain and how it works psychologically. A key pioneer of what we would now recognise as experimental cognitive psychology is the German psychologist Hermann Ebbinghaus. Unlike Wundt and Titchener, he believed that memory was amenable to systematic scientific study, without the need to rely on the limitations imposed by the introspective method. In 1885, he published a landmark paper called “Memory, a Contribution to Experimental Psychology.” This remains today a classic description of the basic principles of experimental psychology and research methods. This included the use of precise control over conditions of the experiment, a tradition shared by others, such as Wundt and Titchener, but also accepting the inevitable impact of extraneous factors impacting on the results. Such measurement error interferes with our ability to accurately test the particular process that we are interested in. For this reason, Ebbinghaus emphasised the need for statistical approaches to estimate the effects that were of interest from the average of the repeated measurements, while at the same time, obtaining estimates of the measurement error. This statistical approach was fundamentally different from what came before. Ebbinghaus applied his method to repeated measurements with himself as the sole subject over prolonged periods. Today, we are much more likely to test repeated research participants. Perhaps the most well-known contribution is the Ebbinghaus forgetting curve. Although it may seem obvious to us, no one had previously measured the time course over which we seem to forget information. Clearly, we can retain some information for years or for our entire life, while other information can be lost to our memory very quickly. In developing his experimental methods, Ebbinghaus reasoned that some information, such as poetry, carries with it a complex set of associations, images, and meanings beyond the words or syllables that make it up. These may allow us to recall a sentence or even a whole verse after a single reading. Such material was not ideal for Ebbinghaus’s purposes. So he chose instead to use a standard set of supposed nonsense syllables, such as “buh” and “pluh” and “kik,” that have no meaning in his language. Before being able to measure forgetting, the set of syllables first had to be learned. And this was done through repetition, until the set could be repeated back 100% accurately, a level of performance known as the criterion. He then measured how much was retained in his memory after a delay that he varied from minutes to several days. To measure retention, he learned the same list of syllables to see how many repetitions it took to get back to the criterion of 100%. The difference between the original and the late learning was known as the savings score. The better retention, that is, the less the forgetting, the fewer repetitions he would need to relearn, and so the larger the savings score. Once done, he started again with a new set of nonsense syllables. When averaged over many individual trials with different delays or attention intervals, the result was the classic Ebbinghaus forgetting curve. This shows that most forgetting happens very quickly in the first minutes and hours. But then, forgetting slowsdown such that there is very little further forgetting of material after four days, even when intervals of a month were measured. Transcripts by 3Playmedia Week © King’s College London 2018 6. Through these precise, repeated experimental methods, Ebbinghaus offered some fundamental insights into the processes of memory that were inaccessible to the method of introspection and paved the way for experimental cognitive psychology of memory that you’ll learn more about next week. Transcripts by 3Playmedia Week © King’s College London 2018 7. Module: Psychological Foundations of Mental Health Week 1 Introduction to cognitive psychology Topic 1 Foundations of cognitive psychology from Plato to Pavlov - Part 3 of 3 Professor Richard Brown Department of Psychology, King’s College London Lecture transcript Slide 3 Until the late 19th to early 20th century, all psychology was effectively cognitive psychology, an attempt to understand the processes and structures that constituted the mind and that permitted the mind to control a range of adaptive functions. However, in the early 20th century, dissatisfaction was growing with the existing approaches, particularly those based on introspection. The reaction against introspection, specifically, and the definition of the psychology of the mind in general finds its voice in the growth of behaviourism. Many of the criticisms of introspection were actually raised by proponents of the method, including Wundt himself. First, it could be unreliable, in that one person’s introspection might be very different from another’s, or even their own introspection on a different occasion. Second, it was unrepresentative. Those doing the introspection were highly trained and therefore their experiences may not be representative of people in general. Third, it was limited in its use. It could not be used with children or those with low intellectual ability or limited language skills. Finally, it was limited in the areas of psychology to which it could be applied. More specifically, by definition, it could not be used to examine structures and processes that were unconscious or happened so rapidly that they were not accessible to the observer to describe. It could also not be applied easily to areas such as language and memory. And for that reason, they were explicitly ignored. This latter criticism also pinpointed a wider issue. Until the late 19th century, psychology was human psychology. The idea of animal psychology did not exist, and indeed, made little sense. While it could be argued that humans have unique faculties and abilities, it is also clear that much of our important behaviour is governed by mental properties that we share with animals, such as the ability to perceive and to react, to learn and remember, to find and approach food, and avoid danger. When the study of animal or comparative psychology started to emerge, there was an obvious Transcripts by 3Playmedia Week © King’s College London 2018 1. danger that people would start to apply human constructs related to the human mind to other animal species, without the ability to confirm their existence empirically. In response to this danger, the English psychologist C. Lloyd Morgan proposed what has come to be known as Morgan’s Canon. This bears reading to understand how important and influential it became. “In no case is an animal activity to be interpreted in terms of higher psychological processes if it can be fairly interpreted in terms of processes which stand lower in the scale of psychological evolution and development.” This set an important constraint on our theories of animal behaviour. We need the simplest possible explanation that explains what can be observed without invoking higher, that is, human, psychological processes. This makes good scientific sense and is consistent with the general scientific principle of parsimony or Occam’s razor. Namely, that when you have two alternative explanations for the same observation or fact, the simplest or one with the fewest assumptions should be selected. While not inevitably correct, it’s a pretty good guiding principle. In relation to animal behaviour, it suggests that we should be looking for psychological explanations that avoid the need to infer human or human-like faculties, such as consciousness or volition or even thought. While obviously important for the study of animal behaviour, Morgan’s Canon also has an important implication for human psychology. If we don’t need human processes to explain behaviour in animals, why do we need such unobservable and hard to study processes to explain our own behaviour? In other words, do we need cognition at all? However, it’s important to note-- Morgan did not deny the idea that internal or cognitive events existed in animals. Instead, he was cautioning against invoking or inferring them as explanations for animal behaviour when simpler explanations existed and that did not require them. Slide 4 Such thinking, and other similar, combined with dissatisfaction with much of 19th century psychological methods and theory, and was instrumental in the rise of a whole new approach to psychology behaviourism-- in American spelling, spelled without the u. Although early influences of behaviourism can be found dating back before the 20th century, including Morgan, the school of psychological theory and practice was officially founded in its modern form by the American Psychologist JB Watson. At a lecture given by him at Columbia University in New York in 1913 called “Psychology as the Behaviourist Views It”, that can be considered the behaviourist manifesto. In his work, he rounded on both the structuralist and the functionalists, rejects out-of-hand the method of introspection, and denies the need to invoke faculties such as consciousness, volition, imagery, and perception. Essentially, he stated that the study of mental states is fundamentally unscientific and needs to be abandoned. We see here just a couple of the uncompromising criticisms that Watson made on the psychology of the day in 1913. Slide 5 He felt it had completely failed in its efforts to be recognised as a natural science and advocated the abandonment of its underpinning principles and methods. For behaviourists, the most fundamental unit was that of association, how the animal or human establish links between two stimuli, or between a stimulus and a response, or between a response and an outcome or adaptive behavioural goal, in other words-- learning. In effect, the birth of behaviourism was the start of what we now call learning theory, an area of psychology that continues to flourish, albeit in the context of a much wider cognitive framework. The radical shift in approach advocated by Watson meant that things that we consider as essentially Transcripts by 3Playmedia Week © King’s College London 2018 2. cognitive, such as human language, reasoning, and problem solving, would become pushed to the margins of psychological study, as proposed by Watson and other behaviourists. This was not to deny their existence. He called them private events, only to say that they must be explainable by simpler, observable processes, or public events. The study of basic associative processes is also central to the neural scientific study of learning and the mechanisms by which neurons alter their patterns of firing based on the relationship between a stimulus and its outcome, such as reward. That’s covered in the Fundamentals of Neural Science module. Slide 6 Let’s try to define behaviourism. It’s an approach to psychology which limits itself to the description of relationships between observable environmental events and ensuing observable behaviour of organisms in the environment. Typically, behaviourism rejects subjective experience as a proper topic of study and resists explanations of observable acts in terms of inferred but unobservable mental processes. The first part is a clear statement of the beliefs that we need to avoid psychological explanations that go beyond the inputs, what can be directly observed, both in terms of the environmental event that surrounds the animal or human immediately before or during the thing that we want to explain, and the outputs, the observable behaviour or response. The second part of the definition here reflects Morgan’s Canon, that we should resist using unobservable or subjective experience in our explanations. In effect, behaviourism stated that the study of unobservable mental processes, Watson’s private events, is fundamentally unscientific and not valid targets for study. If the brain and the mind is a black box, it can and should remain on. The private events should remain private. However, rather than simply limiting study to that which could be observed and ignoring mental processes, later behaviourists sought to apply the principles to explain even the most complex of behaviours, behaviours such as language and social behaviour that would seem to require acceptance of an inner state. Slide 8 A note on behaviourist terminology before we go further. Varied terms are used in behaviourists’ research and explanation. In controlled experiments, the first link or input is often called the stimulus, such as a buzzer or a light. While the term antecedent is used to describe a natural, less-controlled situation, event, or circumstance that precedes a subsequent output. And so it could be deemed to have influenced it. This is the term used in what behaviourists call functional analysis, which we return to later. In terms of the third link or output, the term response is used in controlled experiments, particularly where there is a single simple output, such as a lever press. When the output is more complex and where functional analysis is required, the term behaviour is typically used. Although the language is different, the basic input/output model remains the same. Finally, just to note the term used for the contents and what we have called the black box-- the inner state. This is not a term that behaviourists use, except dismissively. Instead, these unobservable processes are typically referred to as intervening variables, Watson’s private events. They merely provide a link in the chain between the input and the output. Pull the chain and things move at the other end, regardless of the number or nature of the hidden links. We do not need to know in detail what they are or what functions they solve, just that they provide a link between the input and the output. Transcripts by 3Playmedia Week © King’s College London 2018 3. Slide 9 As with all new scientific approaches, they rarely emerge fully formed in the mind of the pioneers, but build on work conducted by others and then brought together in a novel way. For Watson, two key influences that shaped his thinking amongst many were the work of the Russian physiologist, Pavlov, and the American psychologist Edward Thorndike. First, let’s consider the pioneering work of Pavlov. One of his many interests was the physiological reflex. These were behaviours or reactions which did not seem to be learned, but which could be explained by built-in or innate properties of the nervous system. For example, we blink quickly and automatically to protect our eyes if an object comes too close. And we produce saliva in our mouths when appetising food is put on a plate in front of us, helping to prepare our digestion. The eye blink and the salivation reflex are automatic physiological responses to a specific type of stimulus. It was shown by all people and across animal species, typically from birth, and are likely to be unlearnt. They are also extremely difficult to prevent voluntarily. When putting drops in your eye, try not to blink when a drop hits your eyeball. Such reflexes are also preserved in animals and humans with major damage to the central nervous system, something that can be explained by the neuronal pathways required for their control. Pavlov studied various reflex digestive reactions in dogs that had been prepared surgically so that their saliva or gastric enzyme secretions could be collected and measured, work that eventually led to him being awarded the Nobel Prize of Medicine in 1904. Like most people, Pavlov was aware that a dog will start to produce saliva, not just when food was presented when it started to eat, but when it first sees the food, or even hears the cupboard containing the food being opened. For his experiments, Pavlov wanted to measure digestive fluids without getting food in the way of accurate measurement. We only salivate to food when we are hungry. So once the dog had eaten, it could not be studied again for several hours. For these reasons, he systematically trained the dogs to make their reflex digestive secretions without the need for food at all. Slide 10 He did this by a systematic and gradual process of pairing the natural food stimulus, food in the mouth, with a neutral stimulus that would not normally produce salivation, for example, the ticking of a metronome or the ringing of a bell. By repeatedly pairing the food and the bell, an association was formed with the result that the dog produced saliva to the sound of the bell alone without the need to present food. Pavlov called the salivation a psychic secretion. This form of learning has come to be known as classical conditioning and relies on a temporal pairing of two stimuli so that the natural properties of one transfer to the second. It is a form of Stimulus- Stimulus, - or SS associative learning. The terminology of classical conditioning is as follows. The food is known as the natural or Unconditioned Stimulus, or US. And the salivation to food, the natural or Unconditioned Response, or UR. Salivation to food is an unconditional reflex. The bell, on its own to begin with before it is paired with the US, is neutral and produces no response. During the process of pairing of food in the mouth, the US and the bell, the learning or conditioning stage, the dog continues to respond in the presence of food. And so the salivation is still considered a UR. Transcripts by 3Playmedia Week © King’s College London 2018 4. Finally, after repeated pairings, the bell provokes salivation when presented in the absence of food. In other words, the bell has acquired properties of the US and is now a Conditioned Stimulus, or CS. The salivation, because it responds in this context to the bell and not the food, is now called a Conditioned Response, or CR, and represents a new conditional reflex. Interestingly, once the bell has acquired the ability to elicit salivation, the bell can be paired with another neutral stimulus, such as a light, until it too becomes a conditioned stimulus, so-called secondary conditioning. Slide 11 While a useful way to measure physiological reaction in digestion without contamination by food, classical conditioning became a crucial method to study the mechanisms of associative learning, including how learning developed over time or declined. The schematic graph on the left shows the amount of saliva produced as a CR to the bell alone, before conditioning, after five, 10 or 15 pairings. The amount of saliva produced increases with the number of pairings, up to a physiological limit, reflecting the bell acquiring strength as a CS to produce a CR. The conditioned response does not last forever. If the bell no longer serves as a signal for the imminent arrival of food, the CR eventually stops, a phenomenon known as extinction. Once again, the bell becomes a neutral stimulus. However, not all of the learning has gone. It’s simply no longer serves a useful function. If the dog is exposed to the bell, a few hours after extinction, the animal may show a reemergence of the salivation response, although this will be short lived. This is called, spontaneous recovery. This is best explained by the re-activation of a dormant learned association, rather than new learning. The classical conditioning paradigm can be a useful tool to study things other than learning. For example, it can be used to infer the ability of the animal or the human to discriminate between difference stimuli. Animals conditioned to produce a CR to a specific CS will produce the same CR to a similar CS. The strength of the CR will be related to how similar the new CS is to the original. For example, an animal trained to a tuning fork at a specific frequency will show a systematic reduction in saliva production for higher or lower frequencies, the so-called generalisation gradient, as shown in the figure. Slide 12 It is tempting to consider that through the process of classical conditioning, the dog comes to know, to become consciously aware, that the bell signals the imminent arrival of food. However, this was not Pavlov’s contention. As this cartoon suggests, the simple ability of a stimulus to produce a response, however complex, does not necessarily imply that there is a cognitive mediational process, or conscious knowing. Remember Morgan’s Canon? We should avoid invoking cognitions if there are simpler explanations. Similarly, a change in the strength of the CR to a similar CS indicates that the animal has discriminated between the original and the new CS, but does not necessarily imply that discrimination was a conscious one. Slide 13 Perhaps the clearest demonstration that classical conditioning does not require cognitive explanations comes from the observation that even the simplest of animals can be conditioned and learn new stimulus-stimulus pairings. The cute creature shown here is the California sea hare, or sea slug, Aplysia californica, which can grow up to 75 centimetres long and weigh in at up to seven kilogrammes. While cute, it has a very simple central nervous system, made up of only about 10,000 neurons, compare to 100 billion in the human brain. Yet, it still shows the ability to learn through classical conditioning. Transcripts by 3Playmedia Week © King’s College London 2018 5. This is shown in these acquisition and extinction graphs from an experiment that measures the conditioning of the natural reflex, the traction, of the animal’s delicate syphoning gills to touch, much the same as a land snail retracts its eye stalks when touched. The top line shows the animal that received the paired neutral US pairing, showing the growing strength of the neutral stimulus to evoke the response, as it becomes a conditioned stimulus. The other groups had similar stimuli or random pairings of stimuli. Reliable conditioning in the paired group happens within as a few as 15 pairings, and was retained for several days. We then see the identical extinction of the conditioned reflex when exposed to random pairings, exactly as shown by rats, dogs, and humans. As a side note, in demonstrating that we do not need to invoke cognitions in the presence of classical conditioning in such simple animals, neuroscientists have a means or model to examine the exact processes within and between cells that allow learning through classical conditioning to take place, processes that are almost certainly going to apply in all animal species, including humans. For example, in Aplysia, it is possible to study which part of its nervous system might be involved in the learning. The investigators found that the removal of one particular group of neurons abolishes the ability to learn through classical conditioning. One of the co-authors of this classic study, Eric Kandel, won the Nobel Prize for physiology in medicine in 2000 for his work on the physiological basis of memory, much of it using Aplysia. Kandel started out as a psychiatrist and studied psychoanalysis, making him want to understand how memory worked. He was at Harvard University at the same time as Skinner and was undeniably influenced by his work. However, while Skinner remained firmly within the psychological domain, Kandel’s work forged links between psychology and brain science that we recognise today, a long way from his roots in Freudian psychoanalysis. Slide 14 Before leaving classical conditioning for now, we need to note an influential, if highly questionable, study conducted by JB Watson and colleague Rosalie Rayner and published in 1920 in one of the earliest volumes of the new Journal of Experimental Psychology. The study was intended as a demonstration of how basic principles of classical conditioning can explain the development of neurosis and specifically phobias. Fear is a natural reflexive response to naturally occurring stimuli in our environment, something that has developed through evolution as a critical survival tool. As well as the emotional reaction, it has physiological components such as an increased heart rate and breathing, and behavioural components such as avoidance or escape. The phobia-- these natural fear responses occur in the context of environmental stimulus that’s typically perceived as safe and unthreatening to most people. A prevailing psychological explanation at the time came from Sigmund Freud and his psychoanalytic concepts. Famously, he wrote a case study of a young boy called Little Hans who had developed a fear of horses falling over and of being bitten by them. Freud’s analysis, published in 1909, had concluded in essence that the fear was actually related to the Oedipal complex, that Little Hans unconsciously wanted to have sex with his mother and that he was afraid that his father, symbolised by the horse, would cut off his penis. Poor Little Hans had clearly seen horses in the street. And these, for some reason, had caused him to develop a fear of them. For Freud, to use the behaviourist terminology, the first link, the actual experience or stimulus, was less important than the second link, the internal, unobservable, unconscious or private event. Transcripts by 3Playmedia Week © King’s College London 2018 6. Slide 15 Not surprisingly, this gave a perfect opportunity for Watson to show that there was no need to infer these private or internal events, that phobias could develop by the process of conditioning. With that in mind, he set about demonstrating that it was possible to simply and quickly condition fear in a baby using Pavlov’s principles. This was a practical test of a theory which he had published a couple of years before by Watson and a different Morgan, that infancy was the time when patterns of emotional reaction were shaped by experience and associations made to stimuli, which could evoke them. Slide 16 Watson and Rayner reasoned that they could take a neutral object and condition a fear response by pairing it to the stimulus that evoked fear. They chose a white rat as the neutral stimulus. Albert, then about nine months old, showed no evidence of fear when presented with the rat. Indeed, he was curious and touched it as it ran around him. He was also untroubled by a large white dog brought up to him, or a wide range of other stimuli, such as a rabbit or a monkey and various masks. Watson described Albert as generally stolid and unemotional. The next stage involved testing Albert’s response to another stimulus that might provoke an emotional reaction. After trying various things with no obvious effect on Albert, they found that striking a steel bar with a hammer unexpectedly from behind evoked the expected and, for Watson, hoped for response of crying and distress or fear. This was the first observation of emotion that Watson had observed in Albert. Watson has his unconditioned stimulus, or US, and could provoke crying, a sign of fear or distress, as the natural or unconditioned response, the UR. The next stage involved re-presenting the rat to Albert, but this time, at each presentation, pairing it with a distressing sound of the hammer hitting the bar. As before, Albert exhibited signs of distress. This pairing was repeated seven times over a period of a week. Eventually, Watson recorded the following in his notes. “The instant the rat was shown, the baby began to cry. Almost instantly, he turned sharply to the left, fell over on his left side, raised himself on all fours, and began to crawl away so rapidly that he was caught with difficulty before reaching the edge of the table.” These were classical signs of fear. The rat was now a conditioned stimulus. And poor Little Albert developed a conditioned fear response to a previously neutral object. Watson went on to show five days later that this conditioned fear also extended or generalised to other white furry objects, such as a rabbit. When Albert tried to escape, Watson noted this was a most satisfactory test. Other generalisations of the fear response, although less pronounced, were observed with a dog, a pile of white cotton wool, and Watson himself wearing a large fake white beard. The study ended soon after. And it was noted by Watson that contact with Little Albert was lost at that point before they could seek to undo the learned fear by a process of deconditioning. Watson expressed concern that the learned fear would persist indefinitely. Much has been written about the study and little Albert, whose identity and fate remain a mystery. Considerable effort has gone into tracking him down later in life, but none of them conclusive. The study is unquestionably unethical by today’s standards, but it is informative of the thinking of the early behaviourists as it is about the origins of neurosis. However, it also triggered subsequent work in the development of positive therapeutic uses of conditioning principles and was a key step in the Transcripts by 3Playmedia Week © King’s College London 2018 7. evolution of what came to be called behaviour therapy, that we will come back to later in the week. Transcripts by 3Playmedia Week © King’s College London 2018 8. Module: Psychological Foundations of Mental Health Week 1 Introduction to cognitive psychology Topic 2 The heyday of behaviourism: operant learning - Part 1 of 3 Professor Richard Brown Department of Psychology, King’s College London Lecture transcript Slide 4 Let’s look at another major influence on behaviourism, Edward Thorndike. Thorndike was another pioneer of behaviourism, and like Pavlov and Morgan, he studied animals. He did so to understand what might be seen as distinctly human attributes, such as intelligence, consciousness, and the mind. Like others, he viewed that much of this could be explained in terms of reflexes and connections. However, he also thought that in humans what we called our mental life, required the assumption of internal events that lay between them and mediated the relationship between the stimulus and the response. Something that Watson later rejected in his strict behaviourist stance. Mediation is a key concept in cognitive psychology. Indeed, cognitive processes are, by definition, mediational. They underlie an observed relationship between one observable event and another. The debate between behaviourism and cognitive traditions lies in the nature of those mediating events, and whether they are necessary for a comprehensive model of the mind. Slide 6 One of Thorndike’s practical contributions was the development of specific apparatus to measure aspects of animal intelligence. He used the ability to learn as a measure of intelligence, although that narrow definition is not used today, historically challenged by later behaviourists, who rejected the need to invoke such ideas for humans or for animals. Nevertheless, his experiments and apparatus subsequently became refined by later behaviourists. A good example is his puzzle box that he used to study learning in cats. In a typical study, a hungry cat is placed in a cage with a bowl of food out of reach. The cats’ task is simply to escape the box to get the food. The experimenter is interested in how long it takes the cat to get out. Typically, the first time in the box the cat will show a wide range of behaviours, such as howling, reaching through the bars, turning in circles, and so on. Eventually, by chance, the animal will press a lever that opens the cage door allowing it to reach the food. The next time in the cage, the same will happen but, with repeated trials, the time it takes the animal to get out of the cage will become less and less, as the animal learns that pressing the lever opens the door. Thorndike called this trial and error learning and his learning theory was called connectionism. Operant learning differed from classical conditioning in that it was based on the association, the stimulus and a response, and the subsequent outcome or reward, rather than the simple pairing of a US and a CS. For this reason, operant conditioning is sometimes seen as an example of contingency learning, or SR learning, when a specific response is made contingent upon Transcripts by 3Playmedia Week © King’s College London 2018 1. a specific stimulus being present. In contrast, classical conditioning is sometimes called contiguity learning, based on the temporal contiguity or pairing of two stimuli. Various other experiments conducted by Thorndike showed how animals can learn more complex responses or sequences by linking together or training. This provided a means by which animals can learn more complex behaviours by using the same basic building blocks. Slide 7 A crucial element of Thorndike’s theory was the role of the positive or negative effects of the behaviour, and whether it left the animal, what he called, satisfied or dissatisfied. This is stated in his law of effect. Essentially, where a response led to satisfaction, escape from the box and getting the food, that is reward, the same behaviour was more likely to be repeated on the next occasion. While the behaviours that lead to dissatisfaction, that is remaining in the cage and hungry, were correspondingly less likely to be repeated. In this way, the animal, including humans, learned over time based on the consequences of their actions. The emphasis on effect represented a fundamental step in behaviourist thinking. Behaviourism was no longer just looking to explain associations between a stimulus and a response, but considered the critical role of the consequences of that response for the organism and its role in future behaviour. Thorndike’s basic law of learning lies at the core of later learning theories and is commonly called operant or instrumental learning. Rather than simply being a simple two term stimulus response contingency, it is based on a three term, or SRS contingency, the situation, the response and the effect. Slide 9 Now let’s look at a quick example of a three term contingency in action. You are leaving your house and looking up as you leave and see that there are dark clouds gathering. You continue, it rains, and you get wet. The next day as you leave the house, you see that it is cloudy again. You go back and get your umbrella, just in case. Sure enough, it starts to rain but this time you have your umbrella and arrive at your destination nice and dry. The next day, it is sunny, so you don’t bother to take your umbrella. Although seemingly trivial, this illustrates associative learning in action. The presence or absence of clouds in the sky serve as discriminative stimuli, sunny or cloudy. Each of which is associated by reinforcement with a particular consequence, dry or wet, depending on the contingent behaviour: taking your umbrella and using it or leaving it at home. The terms used by Thorndike of situation, response, and effect were later replaced by discriminative stimulus, response and reinforcing. Another terminology for the three term contingency that we will use later is antecedent, behaviour and consequence or ABC. Transcripts by 3Playmedia Week © King’s College London 2018 2. Module: Psychological Foundations of Mental Health Week 1 Introduction to cognitive psychology Topic 2 The heyday of behaviourism: operant learning - Part 2 of 3 Professor Richard Brown Department of Psychology, King’s College London Lecture transcript Slide 3 Building on the foundations of Thorndike and the seminal writing of Watson, behaviourism’s most famous proponent was BF Skinner. More than any other psychologist, Skinner deﬁned and codiﬁed both the conceptual and theoretical framework of behaviourism, standardised the tools to investigate it, and deﬁned its language. Building on Thorndike’s law of effect, he identiﬁed as critical the consequences of behaviour in determining whether that behaviour increased or decreased. In keeping with Watson’s views, he considered that there was no fundamental distinction between the process that shapes the learning and execution of simple acts and the most complex behaviours that early cognitivists ascribed to such things as free will and consciousness. It is worth noting that Skinner was far more than an experimental psychologist. He took the principles of behaviourism and considered their inﬂuence on child development and differences between individuals, on education and the criminal justice system and its wider impact in shaping society and culture, both positively and negatively. In addition to many hundreds of scientiﬁc papers, he authored philosophical works, such as Beyond Freedom and Dignity And Walden Two, a work of ﬁction that imagines a utopia based on the principles of behaviourism. Slide 4 Skinner developed a detailed categorisation system to describe different types of consequences depending on one, whether they increased or decreased behaviour, and two, whether they were positively added or they were something that was taken away. Although all reinforcers strengthen behaviour change, to avoid confusion of terminology, the term “reinforcement” was and is still limited to a consequence that increases future behaviour, while an outcome that decreases a future behaviour is termed a “punisher” or “punishment.” Within each of these two broad categories, he deﬁned both positive and negative instances. We are all familiar with the giving or obtaining of a reinforcer to increase behaviour, such as a child who tidies their room to earn extra pocket money, so-called positive reinforcement, or more commonly a reward. However, a behaviour may also be increased when a negative outcome is removed or avoided as a consequence of that behaviour. For example, the child may be told that they will not be allowed to watch their favourite TV programme if they don’t tidy their room. Here, the threat of being denied something is the stimulus that is removed by the increased behaviour, and can be considered a negative reinforcer. Transcripts by 3Playmedia Week © King’s College London 2018 1. Note that the everyday usage of negative reinforcer is the same as punishment. That is, if the child is prevented from watching TV because they failed to tidy their room, this is an application of a negative reinforcer, because the intention is that the behaviour, room tidying, will increase in future. In a technical sense, they are not being punished. Punishment can also be applied or removed to effect a decrease in behaviour. Although the terminology sounds strange, we can deﬁne punishment as positive or negative, just as we can with reinforcement. Remember, positive and negative here refer to whether the punisher is added or taken away, not whether it is good or bad, pleasant or unpleasant. So a positive punishment is something that is added and leads to a decrease in behaviour, typically something unpleasant, such as denying a privilege in an effort to decrease an unwanted behaviour such as being rude. Finally, we can punish to reduce behaviour by taking away something pleasant, such as ﬁning for breaking a rule or a law. In this example, the everyday use of the ﬁne as punishment is different for the behaviourists. The ﬁne as retribution for a crime is seen as a punishment. However, the behaviourist would only deﬁne it as a punishment if the ﬁne led to a reduction in the offending behaviour in the future. This raises an important point in much of psychology. We always have to be careful how we understand and use terms, such as “reinforcement” and “punishment,” that have both precise technical meanings and also have a similar but often subtly different everyday usage. We will make sure that we give you the deﬁnitions that you need and encourage you to think about those and use the terms more precisely. Slide 6 One more thing to consider is the nature of the reinforcer or punisher. I will use the term “reinforce” here in a broad sense to include both reinforcers and punishers for simplicity. Some things seem to have the ability to inﬂuence and change behaviour on a universal basis, in humans of all ages, and in animals from the simplest to the most complex. These are so-called primary, natural, or unconditioned reinforcers, using the term “reinforcer” here in the broadest sense, as something that changes behaviour. Many of these relate to basic biological needs, driven by evolution and necessary for the survival of the individual or species. These include food, drink, and sex, which the animal will typically approach and work to obtain. These are sometimes called appetitive stimuli, i.e., we have an appetite for them and a biological drive to satisfy that appetite. In contrast, there are consequences, such as pain, loud noise, or bitter tastes, the animal will seek to either avoid or to withdraw from, that is, to escape. These are sometimes called aversive outcomes, and the stimuli associated with them aversive stimuli. Adding or taking away any of these is a powerful determinant of learning and future behaviour in accordance with our deﬁnitions of reinforcers. With most reinforcers, their power depends on the state of the organism at the time. The same food will inﬂuence behaviour more if the animal has been deprived beforehand, that is, it’s hungry, than if presented after it had eaten a large meal, that is, it is satiated. The strength of the reinforcer is not absolute. We can talk of its value to that individual depending on the state at the time. Other less obvious reinforcers can be more species-speciﬁc. For example, for social animals, physical proximity to others of their same species can be considered as a primary reinforcer. Other reinforcers can be considered secondary, or conditioned, reinforcers. These are stimuli that have acquired reinforcing properties by association with another, often primary, reinforcer. Thus, when a light is paired with food, the light itself acquires the ability to initiate approach behaviour and to inﬂuence subsequent behaviour. We saw an example of this in the Little Albert video as the Transcripts by 3Playmedia Week © King’s College London 2018 2. white rat induced an avoidance response after being paired with an aversive stimulus, the loud noise. In animal behaviour studies, secondary reinforcers have similar properties to primary reinforcers, and show the same effects of deprivation and satiety. Slide 7 We humans seemingly have a vast range of things that change our behaviour, and working out whether they are primary or secondary is complex. Some, such as money, would seem to be a classic secondary reinforcer. It has no intrinsic biological or survival value, but we ﬁnd it intensely pleasurable to receive even in small amounts, such as ﬁnding a coin on the ﬂoor, and painful when we lose. We work to earn money and do things to avoid losing it. Some of the reinforcing value of money will come through its association with primary reinforcers, such as food or drink. However, it also seems to have an intrinsic value in itself. We like having money even when we don’t spend it. Also, intriguingly, we never seem to have too much money. Normal satiety processes do not seem to apply. Money is strange in behavioural terms. Other human reinforcers include social factors, such as love and friendship, praise and approval, or indicators of social status, such as a new pair of trainers or a ﬂash car. Others can include such abstract concepts as self-respect, self-worth, or satisfaction. Working out what is inﬂuencing human behaviour in the real world is often a major challenge. It is typically multi-factorial, layered, and complex, with a competing set of drivers, both reinforcers and punishers, both positive and negative. Typically, we are unaware of what drives our own behaviour, something that psychotherapy often aims to help people understand in order to help them change. We will return to this later in the module when we consider the techniques of applied behavioural analysis in the context of behaviour modiﬁcation and behaviour therapy. Meanwhile, it’s interesting to reﬂect about the things in our own lives that inﬂuence our behaviour, things that increase or sustain a particular behaviour, or which decrease it and maintain that decrease. Some may be common and obvious, but others may be particular to you. We will return to this and address it in more detail later in the week. Slide 8 Let’s check our understanding of these technical terms by a quick quiz. Work through the following slides to work out whether we are looking at reinforcement or punishment, and whether it is positive or negative. Remember the deﬁnitions. Think carefully about what is the behaviour that is changing and whether it is increasing or decreasing. Also, is something being applied or taken away, and if so, what? Slide 10 This short quiz shows how we can classify different ways in which an outcome can change behaviour during learning. While it sounds like an easy application of deﬁnition, you may have found the exercise required some thought. For example, was the behaviour in the classroom that changed a decrease in disruptiveness or an increase in attentiveness? One would deﬁne punishment and the other reinforcement. In a similar way, was the imposition of detention the adding of solitude, positive, or the removal of social contact, negative? In practical terms, the outcomes in terms of behaviour may be the same. Also, supposing being detained or caned actually led to an increase in disruptiveness, rather than a decrease, what does that say about the consequence? While the intention of the teacher may have been to punish by withholding social contact or inﬂicting pain, the student’s behaviour may have been reinforced by the subsequent gain in respect or sympathy from their classmates. This makes an important point. We cannot say a priori whether something that is added or taken away is going to increase or decrease behaviour until we have observed its impact. This is a key element of Skinner’s operant theory. Transcripts by 3Playmedia Week © King’s College London 2018 3. Further, when behaviour does change, whether on an individual, a group, or societal level, we have to be careful to check that the things we assume are causing the change are actually the things which do so. The real drivers may be very different. Finally, we need to be aware of unintended consequences, such as the effect of driving up unwanted behaviour when the intention is to reduce it. One example is a health campaign that emphasises the danger or social unacceptability of smoking. These can actually increase the behaviour in groups such as young people who enjoy the sense of risk and rebellion that smoking provides. In later weeks, we will see the importance of this when behavioural principles of reinforcement are used to analyse and understand what causes particular behaviours to emerge and persist, and the techniques of behaviour modiﬁcation treatment designed to change them. In practice, such approaches almost always use reinforcement as the means to change, increasing a desired behaviour as a means to reduce an undesired one, rather than using punishment to achieve the reduction. The same is true in animal training, such as getting a dog to walk to heel, rather than pull on the lead. This is typically achieved by positively rewarding the walking to heel rather than punishing the pulling. Slide 11 Skinner believed that this same principle would also apply to humans. He said a person who has been punished is not less inclined to behave in a given way. At best, he learns how to avoid punishment. Is this always true, particularly in people? What happens when we compare the impact of a positive reward and a positive punishment on behaviour? You may want to reﬂect on this in your own time. There’s been a multitude of studies that have sought to examine the question experimentally across a wide range of settings and different behaviours, from behaviour therapy to education to promotion of healthy living. Both approaches can bring about change, but no consistent pattern emerges on which is more effective. We will not go into detail here. Instead, we will just look at one experiment as an example of an attempt to investigate the question. One of your tasks this week will be to read this paper that you’ll ﬁnd on KEATS and answer the questions posed. This paper will also introduce you to what is known as one factor theory, that suggests that in operant terms, reinforcement and punishment lie at different ends of a single continuum, while two-factor theory suggests that they are distinct and operate on behaviour in fundamentally different ways. Remember, though, the speciﬁc ﬁndings of the study should not be assumed to generalise to all forms of behaviour or to every situation. Transcripts by 3Playmedia Week © King’s College London 2018 4. Module: Psychological Foundations of Mental Health Week 1 Introduction to cognitive psychology Topic 2 The heyday of behaviourism: operant learning - Part 3 of 3 Professor Richard Brown Department of Psychology, King’s College London Lecture transcript Slide 3 Let’s get back to rats and Skinner’s work. Like Thorndyke before him, Skinner was concerned with operant learning. Unlike Thorndyke, who was interested in animals’ seeming ability to solve problems with his puzzle boxes, Skinner reduced the measurements of behaviour to simple parameters-- the speed at which a response was made and repeated, its intensity, and to a lesser extent, it’s duration. The puzzle box was refined to the operant chamber that is still used today in various forms and is commonly called a Skinner box. Skinner’s preferred lab animals were the pigeon, that was well adapted to learn to peck at a light stimulus at a response, and a rat, which could press a lever with its paw. Operant chambers have subsequently been built that can exploit to the behavioural repertoire of just about every animal species. The animal is put in the box only for the time of the experiment. Although designs vary, all boxes have the following common features. There is a means of providing a stimulus, often a siren or a light; a means of the animal making a response, such as a lever that can be pressed; and a means of delivering a reinforcer or punishment, often a small quantity of food delivered or a mild electric shock. Finally, the box is linked to a controller, now a computer, that determines the presentation of the stimuli, controls the delivery of the reward or punishment, and records the animal’s response. Slide 4 One of the legacies of the work of Skinner and other behavioural psychologists has been a detailed understanding of the precise relationship or contingency linking behaviour and the consequent reinforcer or punisher in terms of how and when it is delivered. This work includes the description of a large number of so-called schedules of reinforcement that appear to have universal applicability. They can both describe and predict the behaviour of rats pressing a lever and of human gamblers pouring their money into a slot machine. We will look at a few examples here and some of the simpler schedules used and defined by Skinner. I will use the word reinforcement in general sense to describe any outcome which changes behaviour. The simpler schedule is one where reinforcement occurs every time a response is made, so-called continuous reinforcement schedule or CRF. Transcripts by 3Playmedia Week 1 © King’s College London 2017 1. In such a schedule, the rat would receive a food reward every time it pressed the lever in response to the stimulus. In a human example, we get a cup of coffee dispense every time we put our money into a coffee machine, usually. More commonly, both in laboratory experiments and in the real world, reinforcement occurs only after some responses, not after all-- called intermittent schedules or partial reinforcement schedules. Two main sets of parameters define the main types of schedules. The first parameter is based on the timing or quantity of responses, whether the reinforcer is given after a set number of responses, a ratio schedule, or after a set amount of time, an interval schedule. The second parameter is based on the certainty or predictability that the response will lead to reinforcement, whether the reinforcer is given after a fixed number of responses or amount of time, or after a variable number of responses or time. These parameters can be combined in many different ways, but we will consider the four main ones here. In a fixed ratio or FR schedule, the rat gets a fruit reward after a fixed number of responses. In an FR2 schedule, the reinforcer is delivered every second response. In FR10, every 10th response, and so on. You can see that a continuous reinforcement schedule is, in fact, a fixed ratio or FR1 schedule, where the rat gets a reward every time it presses a lever. In a variable ratio or VR schedule, the reinforcer is given, on average, every nth response, rather than precisely every nth response, as it is in a fixed ratio schedule. So for a VR10 schedule, the reinforcer may be given at the first response on some occasions and the 20th on others. However, with repeated trials, the average number of responses before reinforcement will be set by the schedule of 10. Gambling machines, roulette wheels, and lotteries all work using variable ratios schedules. On a roulette wheel with 36 numbers, we know that we will win over time about once every 36 spins of the wheel. However, the outcome is unpredictable. We might just as often win immediately after our first bet or have to wait for many spins. It is this unpredictability that makes gambling exciting but also potentially addictive. In interval schedules, the reinforcer is given following a response but only after an amount of time has elapsed since the last reinforcer, regardless of how many responses are made. The interval may be fixed or variable. If a coffee machine takes exactly a minute to brew the next cup before it can be dispensed, we will only get our drink by pushing the button once a minute. We don’t get anything apart from frustration by pushing the dispense button repeatedly before that time. This is an example of an FI schedule of reinforcement. Many other types of more complex schedules have been developed, but we will not consider them here. Slide 6 Why are reinforcement schedules important? One reason is that the choice of schedule has a marked effect on how long it takes to learn to make a specific response when presented with a specific stimulus. Second, it influences the intensity and pattern of responding over time. Finally, it affects the persistence of a learned behaviour, even after the reinforcer is withdrawn. The standard way of measuring behaviour from operant boxes is a graph that plots the cumulative number of responses on the vertical axis against time on the horizontal axis. A tick mark on the curve or line shows where a reinforcer is given. The slope of the line indicates the rate of responding, with the steeper the line, the faster the rate. If the animal stops responding, the line becomes horizontal. This figure shows performance on a variable ratio experiment for an animal that has already learned to respond to a stimulus to obtain a reward. The animal is responding at a fairly continuous rate of just under once a second, to get a reinforcer around once a minute. Early in learning, it is usually necessary to have a continuous reinforcement schedule or one with a high probability of Transcripts by 3Playmedia Week 1 © King’s College London 2017 2. reinforcement. This makes sure that the initial trial and error response is reinforced and is likely to be repeated. Once learned, the ratio can be slowly increased across trials so that reinforcement happens less and less often while maintaining responding. This process is an example of shaping. Slide 7 What does behaviour look like under different ratios of response to reinforcement? This first diagram shows the cumulative rate of responding over time under four different fixed ratio schedules-- reinforcement every 70, 120, 180, and 320 responses. The first thing to notice is the characteristic stepped or staircase shape of the record, with bursts of responding showing as the steep part of the trace until the reinforcement is obtained, at which point the response stops, and the trace shows a horizontal line. This same pattern is sitting whether we are measuring the behaviour of a pigeon, a rat, or human for money. It is a basic characteristic of fixed ratio schedules. The duration of the pause and behaviour increases systematically with the ratio. The more response is needed to obtain a reward, the longer the pause after reinforcement before the animal resumes responding. Once started, however, the rate of responding is the same, whatever the ratio. We see a similar pattern for fixed interval schedules, although the characteristic curve is rather more scallop shaped than a sharply defined staircase. Once the animal has learned the interval, they tend to stop or slow down the rate of responding after obtaining the predictably timed reward and start responding more quickly as the duration increases towards the end of the expected interval. Slide 8 What does behaviour look like with variable schedules? This diagram shows the cumulative rate of responding over time and at different variable interval schedules-- reinforcement, on average, every 1, 2, 3, 6, and 10 seconds. The first thing to notice is that unlike the staircase of the fixed schedules, we see almost continuous behaviour, even after the animal has just obtained a reward. This is because there is uncertainty, although it is unlikely another reward may come immediately, so it is adaptive for the animal to keep responding just in case. However, we also see that the rate of responding is slower the longer the meeting interval between reinforcements or the greater the number of responses that have to be made, on average, to get the reward. This makes good adaptive sense. If the reward is food in the wild, then the animal will adjust the amount of energy it expends by varying the rate of responding. This helps to make sure that there is a net gain in energy in over energy out. We don’t have to assume that the animal is doing this consciously. Slide 9 Next, let’s look at what happens when a response stops being reinforced-- if the pigeon no longer gets a reward when it pecks a key. What happens is that eventually the pigeon stops pecking and ignores the stimulus. This is known as extinction. The pattern of extinction is different in fixed and variable schedules. For fixed schedules, the reinforcement is predictable, so it is more obvious when an expected reward is not delivered. However, rather than stopping immediately, we typically see a continuation in behaviour. Indeed, the rate of responding may even increase before eventually stopping. This is sometimes called pre-extinction burst. Some behaviour will continue, but the bouts of behaviour become shorter and the intervals between them longer before stopping altogether. With variable schedules we see behaviour continuing for much longer. This is because there is greater uncertainty about when the next reward is expected. The animal is less able to determine whether it is just an unusually large pause or whether the behaviour is no longer reinforced. The Transcripts by 3Playmedia Week 1 © King’s College London 2017 3. longer the mean time or mean number of responsibles in the variable schedule, the longer the behaviour will continue before it starts to extinguish. Variable ratio schedules show a similar stepped pattern of extinction to fixed ratio schedules, while variable interval schedules show a more continuous slowing of responding before it stops. This figure shows some of the characteristic extinction curves of the different types of schedules. In general, intermittent or partial reinforcement, whether on a fixed or variable schedule, is more resistant to extinction than when an animal has been previously continuously reinforced every response. This is called the partial reinforcement extinction effect. What does it mean that the animal stops responding after the behaviour is no longer reinforced? Does it mean that the animal has forgotten the association between the stimulus and response and subsequent reinforcer? If this was the case, then we would predict that the animal would take as long to relearn the association if the reinforcement contingency was introduced. This was not what we see. Instead, the previous behaviour is quickly reestablished and much more quickly than it took to learn originally. Note, this is very similar to Ebbinghaus’s use of saving scores as an index of learning with his nonsense syllables. So the animal had not forgotten the association. Simply, it was no longer useful or adaptive to respond to a stimulus that did not ever produce a reward. The learned association remained within the animals learned repertoire and was available to be used when the circumstances changed. Slide 10 Before we leave the subjects of reinforcement, conditioning, and extinction for now, fans of the TV show Big Bang Theory may remember an episode in which Sheldon uses operant principles to condition Penny’s behaviour. If you feel like it, take a chance later to review the clip that is found easily on the web. If you do, and once you’ve watched it, you might want to try the quiz questions to test your understanding of what we have just covered in this topic. Transcripts by 3Playmedia Week 1 © King’s College London 2017 4. Module: Psychological Foundations of Mental Health Week 1 Introduction to cognitive psychology Topic 3 The cognitive (r)evolution - Part 1 of 3 Professor Richard Brown Department of Psychology, King’s College London Lecture transcript Slide 4 One of the most influential and important neo-behaviorists was the American psychologist Edward Tolman. Unlike Skinner, who used simple behaviours, such as key pecking or lever pressing to study learning, Tolman was interested in the learning of more complex or purposive behaviour. He used mazes that the animal, typically the rat, had to explore or learn to navigate to find a reward. The slide illustrates some of the many designs of maze that Tolman used in his studies. The important thing about mazes is that the animal has choices to make, such as turning left or right or sometimes continuing straight ahead. This is somewhat different and potentially more cognitive than decisions to press a lever or press a key. However, we should take care to avoid inferring cognitions, unless the evidence strongly points us in that direction. Perhaps Tolman’s most influential work was around the concept of latent learning. This had been described by others earlier in the 20th century. But it was studied most extensively by Tolman and collaborators. Latent, here, means learning that is dormant or concealed. Latent learning is a term applied to learning that seems to occur through exposure to environmental stimuli without any reinforcement. While there is limited change in actual behaviour during the exposure, the animal shows the ability to rapidly learn subsequent reinforced behaviour. This was taken to imply that the animal was learning without reinforcement. We will look at an example experiment in a minute. First, though, we need to note that learning without reinforcement presented a major challenge for strict operant conditioning models. Operant theorists attempts at explanation were often rather elabor

Psychological Foundations of Mental Health Week 1 Introduction to Cognitive Psychology PDF

Document Details

Tags

Related

Summary

Full Transcript

Upgrade to continue