• Overview

    Overview

    A substantial body of experimental evidence has demonstrated that labels have an impact on infant categorisation processes. Yet little is known regarding the nature of the mechanisms by which this effect is achieved. I distinguish two accounts, supervised name-based categorisation and unsupervised feature-based categorisation, and describe a neuro-computational model of infant visual categorisation, based on self-organising maps, that implements the unsupervised feature-based approach. The model successfully reproduces experiments demonstrating the impact of labelling on infant visual categorisation.
    The model predicts that the impact of labels on categorisation is influenced by the perceived similarity and the sequence in which the objects are presented to infants and that the observed behaviour in infants is due to a transient form of learning that might lead to the emergence of hierarchically organised categorical structure. New evidence corroborates these predictions. I argue that early in development, say before 12-months-old, labels need not act as invitations to form categories, nor highlight the commonalities between objects, but may play a more mundane but nevertheless powerful role as additional features that are processed in a similar fashion to other features that characterise objects and object categories.
  • Background

    Nature and Nurture

    Nativists and Empiricists

    Plasticity

    Comparing across Species

    Instinct vs. Learning

    Levels of the Brain

    Is Language Innate

    Learning to Talk

    Left Brain/Right Brain

    Language Impairment

    Cultural Senses

    Improving the Brain

    Intelligence

    Intelligence Quotient

    Types of Intelligence

    Creativity

    Problems

    Nature vs. Nurture

    We are what we eat. No right-minded individual really believes this. I might like to eat a lot of bread but I don’t look like a loaf or even taste like one. Nonetheless, if I don’t eat my daily bread, my body will shrivel and die. I need fuel to survive. Likewise, my brain needs to experience sensory stimuli in order to maintain a healthy state. People who are locked up in sensory isolation chambers quickly begin to hallucinate as the brain tries to create its own source of stimulation. In fact,
    nilsson_rm_photo_of_20_week_fetus
    the brain relies on sensory stimulation for its own development. For example, we know that auditory cortex (that part of the brain that is responsible for processing sounds) develops faster in utero than visual cortex (the part of the brain with primary responsibility for seeing). This precocious development of the hearing part of the brain is probably due to the fact that sounds can pass into the womb and through the fluid protecting the foetus thereby stimulating the brain with sound. The foetus is not exposed to light until birth.

    images-2
    Yet, the structure of our brains and the way that it makes sense of the world is not simply a product of individual experience. The anatomy of the human brain is quite similar for all members of the species. We all have two cerebral hemispheres, a cerebellum, a midbrain, a brainstem, etc. What’s more, the brain seems to work in the same way for most individuals. The brain regions that do the work of seeing, hearing and moving are roughly the same in all of us. This is no accident of nature. The anatomy of the brain and the way that it works is the product of an evolutionary process stretching over tens of millions of years.

    fp_image
    We believe that our brains function as the library of human knowledge. Unfortunately, the most dramatic but convincing demonstration of this fact is the behaviour of brain-damaged patients who cannot function properly. Damage to specific areas of the brain can result in patients losing memories of episodes in their lives. More recently, neuro-imaging studies have demonstrated that different areas of the brain become active depending on what a person is thinking about. Even aspects of practical knowledge, often referred to as muscle memory, depend on the brain for proper function.
    160px-Plato_Silanion_Musei_Capitolini_MC1377
    For millenia, philosophers have debated the extent to which human knowledge is predetermined. Nativists, like Plato, argued that all human knowledge is derived from mathematical universals. These universals are the innate building blocks of everything we know or can know. Although Plato was not privy to scientific theories of genetics, he nevertheless espoused the view that these building blocks of knowledge are inherited from one generation to the next, irrespective of the worldly experiences of individual members of each generation.
    179px-Aristotle_Altemps_Inv8575
    Empiricists, like Aristotle, believed that everything we know is a product of our experience. The extent to which we share knowledge with each other reflects our shared experience of the world. No amount of verbal description will convey the sensational beauty of a rose to a person blind from birth.

    The nature/nurture debate continues to this day, particularly between philosophers, psychologists and linguists who attempt to explain the higher levels of cognitive functioning in humans, like memory, language and perception. Most commentators now agree that the capacity for high level mental processing is inherited:
    286px-EscherichiaColi_NIAID
    Primitive organisms (like bacteria) don’t have this capacity because their genetic endowment does not specify the means to support mental life. However, there is considerable disagreement about what is inherited. Those in the nativist tradition tend to argue that our genetic inheritance provides us very specific knowledge about the world. For example, we come into the world knowing that most objects are solid, that they move around the world in a continuous fashion, and they don’t cease to exist when they disappear out of sight.

    Latter-day empiricists, having conceded that the new-born mind is not a blank slate, argue that brain systems don’t come equipped with knowledge but rather with the capacity to acquire knowledge, often at a phenomenally fast rate. However, the acquisition of knowledge depends on exposure to the right kind of environmental stimulation. On this view, if we grew up in a virtual reality in which apparently solid objects could pass through each other and change position instantaneously, our reasoning about the physical world would be quite different. The debate has changed from whether nature or nurture is the fountain of knowledge, to how much of nature and nurture contribute to the human condition and how they work together to create a unique individual.

    Plasticity

    Although the basic anatomical structure of the human brain is shared by all members of the species, the detailed micro-circuitry varies subtly from one to individual to the next. For example, the precise cortical areas of the brain that are responsible for motor control are not wired up in exactly the same way for everyone. The human brain is quite flexible in the solutions it finds for common everyday problems, like picking up an object or saying a word. Individuals unfortunate enough to lose a finger have to find new ways of manipulating objects. The brain accommodates to these changes by reorganising the
    motorcortex
    connections(synapses) between the neurons responsible for motor control of the fingers. Brain organisation is plastic.

    The plasticity of the human brain (and the brains of many other species) enables it to benefit from learning. The capacity to learn means that individuals can benefit from experience. This is particularly useful if the skill to be acquired is particularly complicated. For example, if you want to learn to talk, you don’t need to know all the words in the language before you start communicating. You can rely on the environment providing you with this information
    Japanese words if you grow up in a Japanese environment and English words if you mother tongue is English. On the other hand, you had better be good at recognising words. And it turns out that humans seem pre-adapted to achieve this, even very young infants. Evolution has found an efficient division of labour between what we can rely on our environment to teach us and what must be built into the organism through our genes.

    Comparing across Species

    Primitive species, like worms, do not benefit very much from learning. They have a very simple
    220px-Bilaterian-plan.svg
    nervous system and can only display a limited repertoire of behaviours, reacting to changes in light or moisture. Experience has little effect on their structure or behaviour. Sometimes, a species might display w
    180px-Natural_Beehive
    hat appears to be a very complex behaviour. Have you ever marvelled at the intricate pattern of the beehive? A perfect design of nature, you may have thought, suggesting a complicated skill has been mastered by the bee. In fact, the form of the beehive is merely the result of the interaction of the bee’s attempt to pack as many honey cells into as small a volume as possible.

    100px-Hexagonal_lattice.svg
    Honey cells start their lives as spherical objects. In the hive these cells are packed together, each bee attempting to create as much space as possible for itself within an individual cell. When honey cells are packed together, the wax walls will undergo deformation. Since each cell contacts just six other cells, the surface tension of the wax and the packing pressure of the bees will force each sphere into a hexagonal shape. The hexagonal shape maximizes the packing of the hive space and the volume of each cell and offers the most economical use of the wax resource. Hence the labour of the bees and the forces of physics act in consortium to produce the hexagonal shape of the beehive. The bee doesn’t need to know anything about hexagons.

    In humans, however, most complex skills must be acquired through a protracted period of development. Evolution has made us relatively helpless at birth so that we can maximise our genetic inheritance and the rich environment in which we live. The adult human is much more than sum of the building blocks of nature and nurture. Just like hydrogen and oxygen combine to produce a compound (water) with properties quite different from its constituent elements, so do nature and nurture combine to yield a product that is not quite predictable form its origins. The complexity of the human brain is the supreme demonstration of what can be achieved through interactive processes, in this case between nature and nurture.

    Instinct vs. Learning

    Learning involves experience of the world. We may be taught facts and skills in an explicit educational process or we may ‘pick up’ things implicitly without conscious awareness, like riding a bike or avoiding situations which are
    images-3
    unpleasant or threatening. In either case, learning involves adjusting synaptic connections between neurons in the brain. As far as we know, knowledge that something is the case or how to do something is retained in our brains as a pattern of connectivity that can give rise to patterns of activity when we access that information. So learning must involve changes in the pathways that control the flow of neural energy.

    Instincts are innate abilities which do not require the intervention of an external source. At least, they are abilities that require minimal triggering by the environment. For example, the new-born foal does not need to learn how to walk. Neuronal pathways are already laid down that provide the basis for the complex motor coordinations underlying its initial attempts to move around. This knowledge how to perform the task of walking has been pre-programmed to manifest itself at birth. Just as learning ‘tunes’ the neurons to make the foal a more efficient walker, so does evolution ‘orchestrate’ the broad sweep of it’s initial capabilities.

    By and large we believe that animals are driven by instinct but humans are shaped by the culture in which they live. Both claims are true but they underestimate the degree to which human behaviour is instinct-driven and animal behaviour is shaped by the environment in which they live. We tend to think of instinct as synonymous with primitive behaviour. In fact, quite complicated abilities in humans can be instinctual, insofar as they don’t seem to be learnt. For example, new-born human infants exhibit an innate preference to look at visual stimuli which resemble the human face. This makes good sense from an evolutionary perspective, because a preference for human faces is likely to enhance an infant’s chances of establishing and maintaining contact with a caregiver, thereby increasing it’s chances of survival. However, the ability to recognise human faces is not entirely determined at birth. Only part of this skill is instinctual. Some of it is learnt.

    clinton
    If new-borns are presented with schematic outlines of faces, they prefer to look at an outline where a few simple features, like the eyes and nose, are in the right position over an outline where these features are jumbled up. But they don’t prefer to look at cartoon faces over real faces. It is as if infants are born with an instinct to look for objects which have some of the basic properties of faces but they need to experience these objects to learn what a face really looks like. Instinct gets the infant looking in the right direction. Learning takes care of the rest.

    This trade-off between instinct and learning is the rule rather than the exception. Very few skills are entirely instinctual except in the most simple organisms. In fact, we are loath to describe instinctual behaviours as skills. Do you regard your knee-jerk reflex as a skill? Likewise, there are no behaviours which are entirely learnt. All new skills rely on existing abilities (some learnt, some instinctual). We can learn to play chess but only because we already understand the concept of a playing a game, of winning and losing, of moving a piece, of turn-taking, etc. Learning is a process of bootstrapping from your current state of knowledge. It is important never to try and teach somebody something they are not ready for. Always teach something that is just moderately novel.

    Back to the Top

    Levels of the Brain

    Neurons are interconnected into identifiable sub-systems called networks. Some networks are very small and localised to discrete regions of the brain. Other networks involve very large numbers of neurons and can be distributed diffusely over wide regions of the brain. Each network often has a particular task to carry out. For example, the hippocampus (picture of the hippocampus here?) is believed to be involved in helping to memorise experiences such as faces or names. We know this because patients who have suffered damage to the hippocampus either through accident or surgery typically have difficulty remembering new things. They can retrieve old memories but have difficulty storing new experiences in long term memory.

    The manner in which the neurons are connected together in the hippocampus makes it particularly well-suited to this task. The network architecture in Figure XXX shows the basic ‘wiring diagram’ underlying the organisation of the hippocampus. The activity in each neuron is passed to other neurons in a reverberating loop of connections, so that the activity of any individual neuron is influenced by the activity of many other neurons. Eventually, this reverberating activity settles into a stable pattern which corresponds to the brain’s identification of a particular memory. The high level of interconnectivity in this network is good for building memory traces because it can combine information from many different sources (taste, smell, sound, vision) to create a composite pattern of activity that is unique to a particular experience. What is more, this type of network can reconstruct a memory from just a fragment of the original material. For example, hearing a person’s voice can help conjure up their name or appearance. This is because there is a network of neurons where many of the characteristics that define a person are moulded together into a single unique experience. Exposure to just one of these characteristics can trigger a memory of the whole person.

    One of the major tasks confronting brain scientists today is to understand how neural networks function. We know that subtle differences in the architecture of a network can have a massive impact on their capacity to perform different tasks. For example, networks which don’t have feedback connections have considerable difficulty telling apart events which occur at different times. Sometimes, this isn’t a problem. You might only want to know that something happened, not when it happened. More often than not, however, timing is critical. Better put the clothes in the washing machine before you switch it on! A common technique used to understand brain networks better is to simulate them on a computer. By modelling bits of the brain on computer, the brain scientist can investigate how artificial neural networks function. The advantage of using the computer is that it is possible to tinker with different parts of the network to see how changing it (or removing them!) can affect its function. This offers insights not realistically achieved by studying real human brains. It is even possible to train artificial networks on the computer and investigate how neural networks develop in infants and young children.

    Is Language Innate?

    Many species have sophisticated communication systems. Bees can dance to indicate the direction and distance of food. Songbirds use their calls to mark out territory or engage in courtship rituals. Monkeys have a wide repertoire of signals that they use to help manage their elaborate social structures and warn of impending danger. Porpoises and whales communicate in a code that we haven’t even managed to crack. Yet, mankind sits at the pinnacle of sophisticated communication. No other species (as far as we know) has a communication system to match the exquisite accuracy of language. By flapping our tongues and making the air vibrate, we can conjure up very specific thoughts in the brains of others. And not just pre-adapted thoughts like “there’s a snake approaching” but also quite unique, useless thoughts like “if you wink your left eye, I’ll hop on my right leg”.

    How are we able to master such a sophisticated communication system? One obvious explanation is that language acquisition in humans takes a relatively long time—a good two to three years in the young of the species. Maybe our complex linguistic abilities are the product of a protracted period of learning. But learning is not the sole prerogative of man. Some birds reared in isolation are capable of learning songs that are not native to their species, if they only ever hear the non-native song. Nevertheless, given the choice of their native song and that of another species, they will acquire their own. Some part of their brain is tuned to listen in for a particular type of tune.

    Investigators of human language acquisition have suggested that our brains are also tuned to learn a particular type of communication system, but one that is far more complex than any song bird. The protracted period of human language acquisition reflects the need to learn. Nevertheless, this learning is based on a complex genetic endowment for language not possessed by any other species on the planet.

    One of the puzzles of language is where it came from. For many other complex features of homo sapiens, like the human eye, it is possible to trace the evolutionary tinkering that resulted in such a complex organism. For language, however, we cannot identify intermediate stages in its evolution. We don’t know of any primitive languages or missing links that might help explain how this gift evolved. Studies of the vocal tract of Neanderthal Man (picture of different vocal tracts here?) suggest that he would have been incapable of producing the range of sounds that characterise human speech. This suggest that language is a relatively recent achievement, probably less than 100,000 years.

    Learning to Talk

    The new-born infant seems completely helpless. In fact, she comes into the world with an impressive set of abilities. For instance, she knows how to recognise a human face, she can tell the difference between speech and other sounds, she can even discriminate between her mother tongue and foreign languages. And yet it will be several years before she is deemed to have mastered her mother tongue. Clearly, there is a lot to learn: Japanese babies need to discover a different set of facts about their language than English babies. The sounds are different, the words are different and the grammar is different. Depending on which language you learn, you may need to pay attention to the order of the words in a sentence to figure out who is doing what to whom (like English), or you may be able to figure it out just on the basis of the endings of the words, irrespective of their position (like Turkish). It’s important to listen to what you hear.

    Nonetheless, babies from all over the world seem equally adept at learning any language. A Japanese baby brought up in England will learn English as easily as an English baby. Although babies come into the world pre-tuned to acquire human language, they are more flexible learners than their caregivers. For example, a Japanese adult finds it very difficult to articulate an /r/ sound differently from an /l/. They also find it difficult to hear the difference, so row and low sound like the same word to a native speaker of Japanese. Not so the Japanese new-born. They are just as proficient as the new-born English infant at discriminating /r/ and /l/. This ability is retained until the age of about 12 months when they begin to behave more like Japanese adults. In fact, new-born humans are sensitive to all the sound distinctions that are found across the world’s languages. Ironically, they know more about some aspects of language than their parents. In order to behave like their parents, they most lose this knowledge. And that is what happens: by the age of 12 months infants are sensitive only to the speech differences which are important in their own language. About the same time as they start producing their first words.

    During the first year of life, infants are not masters of their native tongue. Indeed, the word infant means ‘without language’. To be sure, they are learning a lot about language: They are developing a sensitivity to the sound patterns that characterise their own language and their attempts at controlling the intricate mechanics of speech production is evident in their babbling. This knowledge, however, has not been put to good use. All this changes in the second year of life when children the world over begin to demonstrate their linguistic prowess (cartoon capturing the stages of language development here?). The most obvious demonstration of this ability can be seen in children’s vocabulary development. Early in the second year, they master just a few words like more, look and no, usually to control the flow of goods and services from their caregivers. However, later in the second year (often around 21 months) a dramatic spurt in development occurs. Children suddenly start producing lots of different words, usually names for objects and actions. This development is often referred to as the ‘naming explosion’. From this point onwards, vocabulary develops at the phenomenal rate of around 10 words per day for the next 4 or 5 years of life.

    We don’t really understand why vocabulary development takes off in this sudden fashion. At the same time, children start to combine words to make up miniature sentences. And they don’t just imitate what they hear. They are creative in the language use, sometimes making intelligent mistakes like pointing at a pair of sheep and saying sheeps. This sudden mastery of language may well be the result of brain development during the second year. Indeed, the brain activity in a 15-month-old child that occurs on hearing a word known to her is different to her brain activity on hearing the same word 6 months later when the naming explosion has occurred. Alternatively, the slow but sure progress that has taken place during the first year of life may form the foundations for the later accelerated growth. Perhaps, infants spend the first year or so cracking the language code. Once they crack it, there’s no stopping them. Whatever the explanation, one thing is clear: By the middle of the third year of life, the linguistic capacity of the human child has far outstripped the communicative abilities of any other species on our planet.

    Back to the Top

    Left Brain/Right Brain

    The brain is divided into two hemispheres, left and right. In any individual, the two hemispheres look almost identical and neuro-anatomical investigations reveal that every region on one side of the brain is connected to the equivalent network on the other side. Nevertheless, in the adult brain the two hemispheres are quite different in their functioning. As a general rule of thumb, the left hemisphere takes care of tasks involving symbolic processes (speech, reading, writing and mathematics) while the right hemisphere deals with spatial relations and music (phrenological map of brain function here?). Another way of describing this division of labour is in terms of the distinction between analytic and holistic thinking. This is perhaps best understood in terms of the way the brain processes music. We know that in the general population, music results in preferential activation of the right hemisphere of the brain. However, trained musicians show preferential activation of the left hemisphere. This difference is not due to innate differences between musicians and non-musicians. It is possible to observe the preferred hemisphere for music change during the course of a musician’s training. Musicians appear to listen to music in a different way from the general population. They listen in a more analytic fashion rather than appreciating it holistically. Of course, this doesn’t mean musicians lose their appreciation for music. They just appreciate it in a different way. In general, behaviour that requires analytic intelligence engages the services of the left hemisphere.

    We still do not properly understand why the brain divides task in the way we observe in most people. After all, it could make perfectly good sense to have a region of the brain exclusively devoted to music but allow that part of the brain to change the way it works as a result of experience. Infants who have had the whole of their left-hemispheres removed before the age of 6 months can still grow up with normal language function, even though the left hemisphere is generally considered to be specialised for language. Evidently, the right-hemisphere is flexible enough to take on unaccustomed functions if called upon to do so early in development. Similarly, the left-hemisphere can take over jobs that the right-hemisphere would do in the event of severe, early damage. If damage occurs later in life, however, the chances of complete recovery are much reduced.

    The differences in functioning between the two hemispheres may be due to a fundamental difference in the character of the neural systems in the two regions. These differences may be present at birth and result in different behaviours being recruited by different hemispheres. However, we have no neuro-physiological data that helps us to pinpoint this difference. It is more likely that initially the two hemispheres engage in a competition to carry out specific tasks. For example, we know that language understanding in one-year-olds is distributed across both hemispheres before it become lateralised to the left-hemisphere by the age of two. Subtle initial differences between the two hemispheres may result in the competition for particular tasks (say language or music) being won by a particular side. Once a hemisphere has won one battle, it may be easier for it to win the battle for other tasks that are similar. Hence, once an analytic task (like language) has been assigned to a particular hemisphere (the left), other like-minded tasks follow suit. Before long, an analytic mode of processing can come to dominate an entire hemisphere. What initially started off as a subtle hemispheric difference, can evolve through experience into global differentiation of function. Of course, we still have to discover what the initial subtle difference might be. It has to account for a stable pattern of hemispheric specialisation observed in most individuals.

    Back to the Top

    Language Impairment

    Language impairment comes in many forms: Specific Language Impairment (SLI) where an individual seems to have perfectly normal intellectual abilities but has difficulty with certain grammatical constructions (like getting the ending of verbs right or understanding complicated grammatical constructions such as “The boy at the end of the queue who is a fast runner kissed the girl”): Dyslexia in which there is a difficulty with reading and/or spelling: Aphasia where brain damage (usually in the left hemisphere) results in loss of language function.

    SLI and dyslexia appear to have genetic roots as there is no obvious brain damage or neuro-anatomical abnormality. However, there is much controversy over the source of the problem in these language impairments. Some researchers attribute the impairment to a deficit in a central processing ability such as knowledge of grammar or lack of awareness of the letter to sound rules. Others suggest that the impairment stem from deficits in peripheral processes associated with listening and reading. For example, it has been proposed that SLI results from a deficit in the ability to process sounds quickly. Any such failing of the auditory system would make it difficult to identify certain speech sounds and have the knock on effect of making it difficult to learn the grammar of the language.

    Cultural Senses

    Although individuals come equipped with the same basic sensory apparatus, the way we make use of our senses can vary widely from one culture to the next. For instance, certain African tribes do not interpret line drawings on paper as representations of objects. This is not because there is anything wrong with their vision or lack of ability to see in perspective. Rather their culture does not highlight this particular kind of cognitive interpretation of line drawings. In general, a culture will highlight interpretations of artifacts in their own idiosyncratic way. Culture acts as a filter on our senses, highlighting certain objects and events while passing over others. This does not mean that we actually see things differently. Eskimos don’t see snow any differently to a Londoner. However, our functional interpretation of objects and events, and hence their importance to us, can differ radically from one culture to the next. Eskimos need to be able to identify different types of snow in order to know what equipment he needs for hunting that day. Our senses provide the landscape for our lives. Our culture provides the infrastructure by which we negotiate that landscape.

    Language is one of the most important tools that a culture uses to sharpen and direct our senses. Simply by naming objects and events in the world with nouns and verbs, we draw attention to their importance for us. Of course, the names for objects and actions vary from one language to the next. But so do the objects and actions that a language and culture choose to name. For example, some languages have only two colour words in their vocabularies (such as Dani from Papua New Guinea) while others have eleven (English). Some languages require that you indicate the shape of an object whenever you describe an action involving it (Navaho). Other languages contain a grammatical inflection on the verb to indicate whether an event described by the speaker was witnessed in person rather than merely reported (Turkish).

    Languages vary enormously in terms of the facts they insist on highlighting about the world. Notice that these facts are almost always communicable across languages (otherwise I would not be able to tell you they exist). But because languages insist that their speakers pay attention to certain details, and because those details are often encoded in the language in a concise way, they become more accessible to that speaker. This process is similar to that experienced by anyone who develops an expertise in a skill or domain of scholarship. You learn a specialised vocabulary to go with that skill. Being an expert involves recognising what is essential and what can be ignored. Language provides the members of a culture with framework for deciphering what is important for a society. Learning a language offers the young in a culture an apprenticeship into membership of their society.

    Improving the Brain

    Textbooks in developmental neurobiology will tell you that you are born with just about all the neurons you’ll ever have and that by 12 months of age you’ve established the overwhelming majority of connections (synapses) between those neurons. Thereafter, the development of the central nervous system consists in the fine tuning of existing connections and cell death (we lose millions of neurons every day). This seems like a pretty bleak outlook for improvement! But, of course, we can learn new facts about the world, even learn new languages (albeit imprecisely) and acquire new skills well into adulthood. So evidently there’s a lot to be achieved through fine tuning of existing connections. More recent research has questioned the assumption that the creation of new neurons and the growth of new connections is stunted shortly after birth: New growth can and does occur, under the right conditions, well into adulthood.

    We still need to understand in more depth just how the brain changes in response to experience. We know that human knowledge (practical and theoretical) is represented by the pattern of connections that exist between neurons in the brain. Learning new facts and skills involves changing that pattern of connectivity. One of the most fruitful ways to improve our understanding of this process is to study learning in children. We still know very little about the way in which brain development influences the development of mental abilities in children. Over the decade, researchers have invented ingenious techniques for investigating the basic abilities of infants and young children. We now have a much more precise picture of what they seem to know how to do very early on and what they have to learn. More recently, methods have become available for studying the way that brain activity changes during development. These methods have yet to be fully exploited. But there is enormous potential to discover now how brain development and mental development are tied together. Perhaps when we better understand how this process occurs naturally in young children, we will be better placed to suggest therapeutic methods for individuals with brain damage, or enhance the abilities that normally functioning individuals already possess.

    We are also building ever more sophisticated computer models of the brain using artificial neural networks. It is not science fiction fantasy to suppose that a time will soon arrive when it is possible to introduce silicon based transplants into carbon based brains to facilitate or enhance functioning. We already use cochlea transplants to facilitate hearing in the hearing impaired. The use of artificial neural networks for more central cortical processing is an obvious extension of this trend. Does this mean that our personalities and unique identity will be lost? Almost certainly not. The uniqueness of any individual stems from the enormous complexity of the human brain. We have only begun to scratch the surface of this complexity. It is unlikely that even highly complicated artificial systems will match the complexity of the most complicated machine known to us
    the human braineven by the end of the 21st century.

    Intelligence

    Man, so we are told, has achieved his present evolutionary status because of his intelligence. Prehistoric man was ‘backward’ because he lacked intelligence. We often assume that successful members of society are the intelligent ones, whereas the failures and drop-outs are in some way deficient. What is this desirable quality that makes us what we are and enables us achieve varying degrees of success in the world?

    The term ‘intelligence’ is often used as if it referred to some objective organ of the mind – something definite that people can have more or less of. In fact, intelligence is more like a skill than an object. Just as possession of a muscle does not guarantee its effective use, so must knowledge be carefully analysed and co-ordinated if it is to be used intelligently. Intelligence is not simply the ability to store facts. It requires an intelligent mind to store and use facts appropriately in a particular situation.

    At this level of generality, a definition of intelligence presents no serious problems. But when we try to specify exact behaviours that could be said to be ‘intelligent’ or to define precisely the skills and processes that make up ‘intelligence’ then we immediately run into difficulties. For centuries, philosophers have grappled with the problem of intelligence. A multitude of definitions have been offered: the ability to carry on abstract thinking; the ability to adapt to relatively new situations; the capacity to learn or to profit by experience; or, circularly, the ability to do well in intelligence tests. There is no doubt that some people are more intelligent than others. For example, adults are generally cleverer than children. Present a five year old child with two sticks, A and B, of unequal length and ask him which is the longer. He will correctly answer A. Next, present him with two further sticks, B and C, of unequal length and repeat the question. He correctly answers B. Finally, ask him which of A and C is the longer stick. Chances are that he will answer incorrectly. Yet to the great majority of adults, the correct answer is both necessary and obvious. It reflects an intelligent understanding. But is the development of intelligence just a question of prolonged experience?

    Intelligence Quotient

    Plato believed that by selective breeding it would be possible to produce an elite group of highly intelligent people to govern his ideal state. More recently, some psychologists have suggested that individuals are born with a certain quota of intelligence which sets limits on their potential for intellectual achievement. Furthermore, this intelligence quotient (IQ) is determined by genetic factors, i.e., the IQ of an individual is largely governed by the unique interaction of his parent’s genes. Broadly speaking, this implies that intelligent parents will produce intelligent children while less able parents will produce less able children. Intelligence is viewed as a capacity that is passed down from one generation to the next. As it stands, this claim is indisputable. Our capacity to think and reason is a capacity inherited from our ancestors as much as our ability to walk or to see. On this view, however, the transfer of intelligence from one generation to the next is no mere social process. It is biologically determined. The social environment can only help the individual achieve this innate intellectual potential. It cannot take him beyond this limit.

    Notice that no precise definition of intelligence is being offered here. Normal procedures to evaluate intelligence involve the use of standardised IQ tests that measure various forms of ability. Standardisation implies that the test, which consists of a variety of problems, such as verbal and spatial skills, has been administered to a large number of individuals from a similar socio-economic background. The test is adjusted so that a distribution of scores is obtained with the greatest number of individuals scoring in the mid range of the distribution. Thus, a typical individuals from an ‘average’ white middle-class background can expect to score 100. As we go up or down the scale from this point we will find fewer and fewer individuals.

    It is important to remember that an IQ score is not an absolute measure. Through the test standardisation, the individual’s score is relative to what other individuals, from a broadly similar background, may expect to score. There are two problems with this procedure. First, the test may be inappropriate for a particular individual because his background is very different from those on whom the test was standardised. To take an extreme example, one African tribe is unaccustomed to perspective cues in line drawings. Any test that incorporated these features would be an inappropriate measure of this tribe’s intellectual capacity. For this reason, IQ tests must take into account the cultural background of the person tested. Culture-free tests have been invented for just this purpose. However, a second objection to IQ tests, whether culture-free or culture-specific, is that they are unable to capture the wide variety of skills that individuals possess. Each person has a unique experience of the world, an experience which enables him to develop a unique combination of talents. In an important sense, he is a walking mini-culture. No test is able to capture this kaleidoscopic variety. Consequently, no test can tap and hence measure the variety of intellectual skills that an individual brings to the testing situation. The American linguist, William Labov, found precisely this limitation of conventional tests when attempting to assess the verbal skills of black children in New York. However, he found that if he got to know the children, allowed them to bring a friend along and sat down on the floor to talk to them, then the black child from Harlem shows that he is a sophisticated speaker from a background where verbal skills are highly valued.

    The IQ test can at best give us a crude indication of an individual’s intellectual skills that are held to be of value by a given culture (or psychologist). At worst, the IQ test provides a gross distortion of the individual’s skills, completely ignoring the complex abilities he has almost certainly developed for other tasks.

    Those who believe that intelligence is essentially a biologically inherited capacity often point to the significant correlations found between a child’s IQ and that of his parents. How could this correlation come about if intelligence is not inherited? All parents provide unique environments for rearing their children. Normally, the environment will be sufficiently rich for the child to exercise a wide range of skills; verbal, spatial, perceptual, motor, etc. At the same time, this environment places limits on the child’s opportunity for growth. Even an innate skill requires a suitable environment for it to function adequately. To a considerable extent, the constraints on the child’s opportunities are associated with the parent – the parent’s job, the family’s economic position, the degree to which they play with the child, the parent’s attitude as to what is important in child development. All these are crucial social factors in determining the development of the child’s intellectual skills. Furthermore, these social determinants have been acquired by the parents themselves in their own development. Many sociological studies repeatedly point to reproduction of life-styles within families from one generation to the next, even to the extent that father and son are often found in similar types of employment. It is small wonder that children fare equally well as their parents when asked to perform an IQ test. The attitudes and opportunities for the development of skills are likely to be very similar.

    Some of the most convincing evidence for the heritability of intelligence concerns the study of identical twins (i.e., individuals possessing identical genetic inheritance). Typical estimates suggest that 80% of an individual’s intellectual capacity is determined by heredity while only 20% of the variation between individuals is determined by the environment. On this basis, identical twins should exhibit an overwhelming similarity in intellectual skills irrespective of their background. In particular, identical twins reared apart should show greater similarity in intelligence than non-identical twins reared together, given the overwhelming influence of inheritance as compared to environment. In a lengthy series of studies, Sir Cyril Burt put forward data supporting precisely this position. More recently, however, Leon Kamin, in a careful reanalysis of Burt’s results, revealed a number of contradictions in the data suggesting that the investigation had not been as objective as one expects of scientific enquiry. This is, of course, not sufficient to discredit all such evidence, but it does indicate a tendency to stress one point of view, without giving adequate weight to the alternative factors. The main criticism of the twin studies is that investigators paid insufficient attention to the phenomenon of selective placement, i.e., the tendency of adoption or fostering agencies to put separated twins in homes of similar quality, thereby diminishing the range of environmental effects.

    Back to the Top

    Types of Intelligence

    The measurement of intelligence is always evaluated with respect to overall performance on a variety of tasks. The most important of these include spatial skills, verbal ability, numerical reasoning, perceptual analysis and memory. The overall measure of an individual’s intelligence reflects their achievement across the board. Often, psychologists report a general measure of intelligence, or g-factor, which underlies performance on mental tests of all kinds. Performance on specific factors of intelligence, such as numerical and verbal abilities, are often reported separately. Tests of verbal intelligence may involve assessing how quickly an individual can name a picture or solve verbal reasoning problems like:
    Seed is to plant as egg is to tree root pollen oats potato bird 1. 2. 3. 4. 5. 6.
    Tests of numerical abilities may assess how quickly and accurately an individual can perform mental arithmetic or solve numerical reasoning problems like: In the following series, the fifth member is omitted. Which is it? 56 35 20 10 … 1 2 3 4 5 6 7

    An individual may show considerable variation in performance levels across these different factors. For example, the same individual may have high spatial abilities but show below average performance verbally. Some of this variation may be systematic and predictable, such as women tending to have higher verbal IQ than men and men tending to have higher spatial IQ than women. (It should be noted that although systematic differences between men and women exist on these measures, they are small differences in average performance across groups of men and women. The variation in performance found within a group, say men, is much larger than the small difference in the averages between the sexes.)

    Psychologists have also entertained the idea of different types of intelligence. For example, Howard Gardner has proposed the theory of multiple intelligences which assumes that intelligence is composed of discrete modules. Each module deals with specific, but different kinds of information which individuals encounter in the course of their regular activities. Among these different modules Gardner includes linguistic intelligence; musical intelligence; logico-mathematical intelligence; spatial intelligence; bodily kinaesthetic intelligence; and personal intelligences (access to personal feelings, relations with other, etc). He suggests that these intelligence modules are genetically pre-programmed, though subject to cultural specialisation and educational assistance. Evidence regarding the functional specialisation of the brain would seem to support this theory of multiple intelligences. Furthermore, the fact that brain damage can lead to selective impairment of particular skills and the fact that there exists individuals exceptional in a particular intelligence, adds credence to the view that intelligence should be viewed as a distributed package of abilities rather than a single general factor.

    Of course, this modularisation still leaves us with the problem of defining the different types of intelligence, and even of defining what constitutes intelligent behaviour. As I sit here at my personal computer, I make frequent typing errors which my word processor often automatically corrects for me. Do we want to count this as intelligent behaviour, even though it is completely deterministic, driven by a computer programme? Many commentators would argue that the relatively simple procedures need for a ‘spell check’ grossly underestimate what is involved in human intelligence. How complicated does the computer’s behaviour need to be before we attribute it with an artificial intelligence? In fact, today’s supercomputers are capable of exceedingly complex tasks. They are far better calculators than humans; they can beat the world champion at chess; small hand-held devices are capable of recognising their owner’s speech; they can build cars; they can navigate spaceships and terrestial vehicles; they can recognise faces and fingerprints; they have even been used as psychotherapists!
    Yet most of us would be loath to equate artificial intelligence with human intelligence. One good reason for ‘looking down’ on today’s machines is that although they may be good at what they do, that’s all they’re good at. They lack the capacity to coordinate multiple artificial intelligences in new and creative ways. Of course, as we progress in our understanding of how humans coordinate their intelligences to solve difficult problems, it is likely that we will be able to exploit this knowledge to build truly intelligent machines.

    Back to the Top

    Creativity

    The most obvious examples of human creativity are to be found in the realm of scientific achievement, literature and art. When we think of creative individuals, we think of people who create new objects or ideas, or situate old objects or ideas in novel contexts. Questioned about the source of their creativity, artists and scientists usually come up blank. The philosopher of science, Karl Popper, had very precise ideas as to how to judge the worthiness of scientific hypotheses but was hopelessly vague about where they came from! One sculptor may see a beautiful angel coming ‘out of the stone’ while another might discover a wolf in the same piece of rock. What is the source of human creativity?

    In fact, most of us are creative every time we open our mouths. The American linguist, Noam Chomsky, pointed out many years ago that most of the things we say have never been uttered before. Consider each of the sentences on this page. I can guarantee that you won’t find another page with the exact same sequences of words in any book ever written by a human being. I offer this guarantee not in testimony of my own creativity but as an example of what each and every one us accomplishes when we speak, the creation of novel expressions.

    Chomsky attributed this creative talent for language to our mastery of a system of grammatical rules. These rules govern how words can be put together to make up sentences. We rarely break the rules even though we may not be aware of them. It may seem strange to argue that creativity emerges from obeying rules. However, it was Chomsky’s great insight to realise that grammar provides the foundations or framework on which to build new expressions. All artists and scientists need tools to express their creativity. The layperson uses grammar as a tool for creative expression.

    Pasted Graphic
    Even simple acts of perception involve creativity. Consider the line drawing. The figure is called a Necker Cube. If you stare at it long enough, you will discover that its orientation shifts spontaneously. The shift results from your own creative perceptual reorganisation of the sensory stimulation of your retina. Of course, you can’t create any interpretation of the line drawing. It has to be consistent with your knowledge of 2-D representations of 3-D objects. Again, we see an internal knowledge base at work here. But given that you know how objects can be represented on a page, the artist can exploit your own creativity to produce delightful representations of ambiguous figures.

    The creative brain achieves novel results by combining well-known building blocks (like words) in new ways. We all do this continuously in the course of our everyday activities. But what marks out true creativity? Why are some creative acts regarded more highly than others. After all, we don’t get excited when someone utters a sentence never heard before. We probably don’t even recognise that it happened. Of course, there is a lot of individual variation in what people regard as creativity. Just consider the controversy caused by some works of art displayed in international galleries. And what is regarded as creative varies with the historical context. The new entrance to the Louvre was an act of creative genius which would not have been so regarded after a few dynasties of Pharaohs. Nevertheless, the source of all creativity, both recognised and overlooked, is probably our capacity to combine familiar objects and ideas in novel ways, a capacity demonstrated from infancy through adulthood. Why an individual comes up with a specific novel combination at a particular point in time that is particularly valued by the culture remains somewhat mysterious. At this stage in our understanding, it’s more appropriate to ask how rather than why.

    Back to the Top

    Problems

    Damage to the brain, as a result of head injury, stroke, or disease can lead to a wide variety of intellectual, social, emotional and motor control problems for the unfortunate individual. The nature of the problem incurred depends on the location of the damage. Different parts of the brain are responsible for different mental processes. Some brain lesions will result in very noticeable problems such as complete inability to talk or to recognise people. Other lesions lead to much more subtle effects such as problems with reading words that are spelt irregularly, like yatch, while regularly spelt words, like pot, seem to be read normally (a condition called surface dyslexia). More often than not, the consequences of brain damage are quite complicated because brain lesions do not fit neatly into the natural lines of organisation of the brain. Lesions do not respect neuro-anatomical boundaries and can have quite diverse effects on behaviour. For this reason, the specific pattern of problems that arises from brain damage is usually unique to the individual involved.

    Unfortunately, there are a wide variety of problems effecting mental processes that can arise even in the absence of post-natal brain damage or trauma. For example, conditions such as autism, Asperger’s syndrome, Down’s syndrome, William’s syndrome and specific language impairment all occur in individuals who have not suffered a post-natal lesion. As far as we know, these conditions are genetically determined though the precise mechanisms leading from the genetic abnormality to the mental deficiency are not well-understood. For some conditions, like autism, it is often possible to identify neuro-anatomical abnormities of the brain, such as irregularities in the structure of the cerebellum, hypocampus and limbic systems. However, this does not necessarily tell us exactly what the cause of the problem is. Autistic children show normal patterns of development in the early years but then their linguistic and social development seems to slow down or even stop. They also engage in stereotypic self-occupied movements. Most, autistic children have an IQ below the normal range even though some individuals may exhibit exceptional skills. For example, there is a very well documented case of the child Nadia who demonstrated a superb drawing talent but who was, nevertheless, severely autistic. In later years, stereotypic behaviour tends to disappear but their intellectual and linguistic skills remain below the normal range.

    For other conditions, like specific language impairment, which is thought to effect over 5% of the population, it has not been possible to identify obvious abnormalities in the brain. Specifically language impaired (SLI) individuals can exhibit a variable range of linguistic deficits but are generally regarded to be within the normal range in non-verbal IQ. For example, it has been suggested that some individuals suffering from SLI have difficulty understanding who the reflexive pronoun ‘himself’ refers to in the sentence ‘John says Peter is scratching himself’, even though they understand the sentence ‘Peter is scratching himself’. This points to a fundamental problem in the operation of SLI grammar. Other researchers have pointed to an inadequacy of auditory processing in SLI, suggesting that the problem may be more do with peripheral speech processing than a central cognitive deficit. However, failure to uncover abnormalities in the brain areas that might be responsible for these types of disfunction make it difficult to evaluate these explanations. The new techniques of brain imaging made available through PET, fMRI and MEG may help us to understand the nature of the problem better.

    Individuals suffering from Down’s syndrome, show physical abnormalities, physical problems like heart disease, poor motor control, and severe mental retardation. In syndromes like these, including William’s and Asperger’s syndrome, researchers have been able to identify the genetic abnormalities associated with the deficit. Nevertheless, the causal mechanisms whereby the genetic abnormality leads the observed intellectual deficits is still not known.

    Back to the Top
  • Names or Features
  • Perceptual Load
  • Conclusions
  • Button Title
  • Button Title
  • Button Title