Cognitive Development and Cognitive Neuroscience

The Learning Brain

Usha Goswami

• Publisher: Routledge
• Year: 2019
• ISBN:
  9781317410041
• Edition: 2
• Topic: Pedagogía
• Subtopic: Psicología educativa

Cognitive Development and Cognitive Neuroscience: The Learning Brain is a thoroughly revised edition of the bestselling Cognitive Development. The new edition of this full-colour textbook has been updated with the latest research in cognitive neuroscience, going beyond Piaget and traditional theories to demonstrate how emerging data from the brain sciences require a new theoretical framework for teaching cognitive development, based on learning.

Building on the framework for teaching cognitive development presented in the first edition, Goswami shows how different cognitive domains such as language, causal reasoning and theory of mind may emerge from automatic neural perceptual processes. Cognitive Neuroscience and Cognitive Development integrates principles and data from cognitive science, neuroscience, computer modelling and studies of non-human animals into a model that transforms the study of cognitive development to produce both a key introductory text and a book which encourages the reader to move beyond the superficial and gain a deeper understanding of the subject matter.

Cognitive Development and Cognitive Neuroscience is essential for students of developmental and cognitive psychology, education, language and the learning sciences. It will also be of interest to anyone training to work with children.

Chapter 1

Contents

Memory

Perception and attention

The perceptual structure of the visual world

Cognitive neuroscience and object processing in infancy

Summary

1

Infancy: The physical world 1

KEY TERMS

Naïve physics

An intuitive understanding we have about objects in the physical world; e.g., that objects that are dropped will fall, solid objects cannot pass through other solid objects, etc.

Full object concept

An understanding that objects are enduring entities that continue to exist when out of sight, or otherwise unavailable to our senses.

Associative learning

The ability to make connections between events that are reliably associated.

What kinds of knowledge are central to human cognitive development? One proposal is that knowledge about the physical world of objects and events; knowledge about social cognition, self and agency; and knowledge about the kinds of things in the world, or conceptual knowledge, are the ‘foundational’ domains for cognitive development (Wellman & Gelman, 1998). These domains could be described as naïve physics, naïve psychology and naïve biology. Infants need to learn about objects and the physical laws governing their interactions, they need to learn and understand social cognition (interpreting and predicting people’s behaviour on the basis of psychological causation), and they need to learn about the kinds of ‘stuff’ in the world (such as animate versus inanimate entities). Clearly, cognitive development in these foundational domains is also dependent on the development of perception, memory, attention, learning and reasoning. Most areas of cognition involve all of these processes at once.

It was once thought that young infants, who are immobile and whose perceptual abilities are still developing, have very limited cognition. For example, for many years it was thought that infants did not develop a full object concept until around 18 months of age (Piaget, 1954; a full object concept requires an understanding that objects are enduring entities that continue to exist when out of view). This seems to be far from the case. Recent work in infant perception demonstrates that a remarkable amount of information about the foundational domains is given simply by watching things happen in the world. As argued in the Foreword, this sensory information and the neural mechanisms used by the brain to process it are probably key sources of early cognitive development. The automatic processing and storage of spatio-temporal information about the world gives the infant emergent knowledge and experience-based expectations about how people and about objects behave, which we can call emergent cognitive frameworks.

These emergent cognitive frameworks are also supplemented by information gained through direct action. Much richer information becomes available when the infant becomes able to reach, grasp, sit and move. Usually, the ability to grasp objects unaided begins at around four months. Now infants can experience making a range of different actions on objects that they can hold, such as toys or their bottle of milk. Another important milestone is being able to sit up without support. Babies usually become ‘self-sitters’ between four and six months of age. Babies who can sit upright by themselves can expand their range of actions on the world of objects. For example, becoming a ‘self-sitter’ has been shown to be related to babies’ understanding of objects as three-dimensional. Finally, the ability to crawl (or bottom shuffle) usually onsets at around nine months. Now babies can choose to move towards objects of interest and explore them further, and can experience different viewpoints. This enables the development of ‘allocentric’ spatial frameworks or mental maps – an understanding of space based on salient landmarks in the environment rather than on the infant’s own position in space.

There are a range of types of learning mechanism regarding the physical world that are functioning from very early in development and that support the emergence of early cognitive frameworks. One is associative learning. Babies appear to be able to make connections between associated events even while in the womb. Once outside the womb, they appear able to track statistical dependencies in the world, such as conditional probabilities between events. Statistical conditional learning turns out to be a very powerful mechanism, as we will see. Another type of learning that appears to be available early is learning by imitation. This may be particularly important for the development of social cognition. Learning by imitation is considered further in later chapters. Finally, infants appear able to connect causes and effects by using what machine learning theory calls ‘explanation based’ learning. This is a form of the ‘causal bias’ that was discussed in the Foreword. The causal inferences made apparently automatically by the brains of infants provide an extremely powerful mechanism for learning about the world. Infant’s brains are not simply detecting causal regularities, but appear inherently to record spatio-temporal information about causal trajectories, perceptual information that includes causal structure. This eventually enables the construction of causal explanations for new phenomena on the basis of prior (statistically-based) knowledge. One cognitive mechanism that is used to help construct explanatory frameworks is learning by analogy. This type of learning is also considered further in later chapters.

Once an infant is able to reach out and grasp objects, much richer information becomes available to him or her.

Memory

Memory is a good place to begin to study infant perception and cognition. After all, without some form of memory, infants would live in a constant world of the ‘here and now’. In order to remember, babies must learn what is familiar.

Memory for objects

Infant memory was originally investigated using rather mundane objects and events. For example, Bushnell, McCutcheon, Sinclair and Tweedie (1984) studied infants’ memory for pictures of simple shapes such as red triangles and blue crosses, which were mounted on wooden paddles. The infants were aged three and seven weeks. Memory for a simple stimulus such as a yellow circle was first developed by asking the infants’ mothers to present the stimulus daily for a two-week period. The mothers were encouraged to show their babies the stimulus ‘actively’ for two 15-minute sessions per day. The babies were then visited at home by an experimenter, who showed them the familiar or habituating stimulus, and also a random selection of other stimuli, varying either colour, shape, or colour and shape. The aim was to test the infants’ memories for these different aspects of the stimuli. For example, to test colour memory, the baby might be shown a red circle rather than a yellow circle. To test memory for shape, the baby might be shown a yellow square instead of a yellow circle, and so on. Bushnell et al. found that the infants retained information about every aspect of the stimuli that they had been shown – shape, colour and size.

Cornell (1979) used pictures of groups of such stimuli to study recognition memory in infants aged from five to six months. In addition to pictures of patterns of geometric forms (see Figure 1.1), he also used photographs of human faces. The babies were first shown two identical pictures from Set 1 side by side, followed by two identical pictures from Set 2, followed by two identical pictures from Set 3, and were allowed to study each set for a period of up to 20 seconds. Two days later they were shown the pictures again, first in a brief ‘reminder’ phase in which each previously studied picture was presented on its own, and then for a recognition phase in which the familiar picture from each set was paired with an unfamiliar picture from the same set. Recognition memory was assumed if the infants devoted more looking time to the novel picture in each pair.

Figure 1.1The stimulus sets used by Cornell (1979) to study recognition memory in infants. Copyright © 1979 Elsevier. Reproduced with permission.

Cornell found a novelty preference across all the sets of stimuli that he used. Even though two days had passed since the infants saw the pictures, they remembered those that were familiar and thus preferred to look at the novel pictures in the recognition phase of the experiment. Their recognition memory was not due to the brief reminder cue, as a control group who received the ‘reminder’ phase of the experiment without the initial study phase did not show a novelty response during the recognition test. Given that the stimuli were fairly abstract (except for the faces) and were presented for a relatively short period of time in the initial study phase, their retention over a two-day period is good evidence for well-developed recognition memory in young infants.

Working memory in infancy

KEY TERM

Working memory

The memory system that temporarily keeps in memory information just received that may be processed for further use.

The capacity to retain information over short periods of time is often called ‘short-term memory’ or ‘working memory’. An influential model of memory in adult cognition is Baddeley and Hitch’s (1974) model, which distinguishes a short-term from a long-term system. The short-term system, called working memory, is thought to enable the temporary maintenance of information while it is processed for further use (e.g., in reasoning or in learning). Working memory is thought to have both visuo-spatial and sound-based (phonological) subsystems, which maintain visual versus auditory information respectively.

Working memory abilities in babies were first studied by Rose and her colleagues (Rose, Feldman & Jankowski, 2001). Rose et al. (2001) measured how many items could be held in mind by infants as they developed, testing the same babies when they were aged five, seven and 12 months respectively. The infants were shown colourful toy-like stimuli, in sets of either one, two, three or four items. Once a particular set had been presented, recognition memory was tested by pairing each individual item with a novel item. Working memory capacity was measured by seeing how many objects the babies recognised as novel. For example, if a baby had been shown a set of four items, but only seemed to recognise two of them in the subsequent novelty preference pairings, memory span was assumed to be two items. Primacy and recency effects were also studied. In adult working memory experiments, participants find it easier to remember the first item of a set (primacy), and also to remember the last item (recency). The question was whether babies would show the same effects.

Rose et al. (2001) reported that memory span increased with age. When they were aged five and seven months, rather few babies could hold three or four items in working memory simultaneously (only around 25% of the sample achieved this span). By 12 months, almost half of the babies had a working memory span of three to four items. Recency effects were found at all ages tested – the babies showed better recall for the final item in the set. Primacy effects were not reported, but have been reported in seven-month-old infants by Cornell and Bergstrom (1983). Hence the working memory system of young infants appears to operate in a similar way to that of adults. Primacy and recency effects in adults are explained in terms of the extra cues to recall provided by being the first or the last item in the list. The development of working memory in infancy has also recently been studied using EEG. Many of these studies utilise delayed response paradigms (such as the A not B finding task) and examples are discussed in Chapter 2.

Memory for events

Some striking studies carried out by Clifton and her colleagues have shown that six-month-olds can also retain memories for events, and do so over very long time periods. For example, in one of Clifton’s studies, 6.5-month-olds were able to retain a memory of a single event that had occurred once until they were 2.5 years of age (Perris, Myers & Clifton, 1990).

Perris et al. (1990) demonstrated this by bringing some infants who had taken part in an experiment in their laboratory as six-month-olds back to the laboratory at age 2.5 and retesting them. During the infancy experiment the babies had been required to reach both in the dark and in the light for a Big Bird finger puppet that made a rattle noise (the experiment was about the localisation of sounds). The reaching session had taken about 20 minutes. Two years later, the children were brought back to the same laboratory room and met the same female experimenter, who said that they would play some games. She showed them five plastic toys, including the Big Bird puppet, and asked which toy they thought would be part of the game. She then told them that Big Bird made a sound in the game, and asked them to guess which one it was out of a rattle noise, a bell and a clicker. Finally, the children played a game in the dark, which was to reach accurately to one of five possible locations for the sounding puppet. After five uninstructed dark trials, during which no instructions about what to do were given, the children were given five more trials in which they were told to “catch the noisy Big Bird in the dark”. A group of control children who had not experienced the procedure as infants were also tested.

KEY TERM

Implicit recall

Recall of information that is not explicitly available.

Perris et al. found that the experimental group showed little explicit recall of their experiences as infants. They were no more likely than the control group to select Big Bird as the toy who would be part of the game, or to choose the rattle noise over the bell and the clicker. However, they showed a clear degree of implicit recall, as measured by their behaviour during the game in the dark. They were more likely to reach out towards the sound than the children in the control group in the first five trials, and they also reached more accurately. If they were given a reminder of their early experience, by hearing the sound of the rattle for three seconds half an hour before the test in the dark, then they were especially likely to show the reaching behaviour. Again, this was not true of the control group. Finally, the children who had experienced the auditory localisation task as infants were much less likely to become distressed by the darkness during the testing than the children who had not experienced the auditory localisation task as infants. Nine of the la...