I have long been interested in classifications of people, in how they affect the people classified, and how the affects on the people in turn change the classifications. We think of many kinds of people as objects of scientific inquiry. Sometimes to control them, as prostitutes, sometimes to help them, as potential suicides. Sometimes to organise and help, but at the same time keep ourselves safe, as the poor or the homeless. Sometimes to change them for their own good and the good of the public, as the obese. Sometimes just to admire, to understand, to encourage and perhaps even to emulate, as (sometimes) geniuses. We think of these kinds of people as definite classes defined by definite properties. As we get to know more about these properties, we will be able to control, help, change, or emulate them better. But it's not quite like that. They are moving targets because our investigations interact with them, and change them. And since they are changed, they are not quite the same kind of people as before. The target has moved. I call this the 'looping effect'. Sometimes, our sciences create kinds of people that in a certain sense did not exist before. I call this 'making up people'.

What sciences? The ones I shall call the human sciences, which, thus understood, include many social sciences, psychology, psychiatry and, speaking loosely, a good deal of clinical medicine. I am only pointing, for not only is my definition vague, but specific sciences should never be defined except for administrative and educational purposes. Living sciences are always crossing borders and borrowing from each other.

The engines used in these sciences are engines of discovery but also engines for making up people. Statistical analysis of classes of people is a fundamental engine. We constantly try to medicalise: doctors tried to medicalise suicide as early as the 1830s. The brains of suicides were dissected to find the hidden cause. More generally, we try to biologise, to recognise a biological foundation for the problems that beset a class of people. More recently, we have hoped to geneticise as much as possible. Thus obesity, once regarded as a problem of incontinence, or weakness of the will, becomes the province of medicine, then of biology, and at present we search for inherited genetic tendencies. A similar story can be told in the search for the criminal personality.

These reflections on the classification of people are a species of nominalism. But traditional nominalism is static. Mine is dynamic; I am interested in how names interact with the named. The first dynamic nominalist may have been Nietzsche. An aphorism in The Gay Science begins: 'There is something that causes me the greatest difficulty, and continues to do so without relief: unspeakably more depends on what things are called than on what they are.' It ends: 'Creating new names and assessments and apparent truths is enough to create new 'things'.' Making up people would be a special case of this phenomenon.

Around 1970, there arose a few paradigm cases of strange behaviour similar to phenomena discussed a century earlier and largely forgotten. A few psychiatrists began to diagnose multiple personality. It was rather sensational. More and more unhappy people started manifesting these symptoms. At first they had the symptoms they were expected to have, but then they became more and more bizarre. First, a person had two or three personalities. Within a decade the mean number was 17. This fed back into the diagnoses, and became part of the standard set of symptoms. It became part of the therapy to elicit more and more alters. Psychiatrists cast around for causes, and created a primitive, easily understood pseudo-Freudian aetiology of early sexual abuse, coupled with repressed memories. Knowing this was the cause, the patients obligingly retrieved the memories. More than that, this became a way to be a person. In 1986, I wrote that there could never be 'split' bars, analogous to gay bars. In 1991 I went to my first split bar.

This story can be placed in a five-part framework. We have (a) a classification, multiple personality, associated with what at the time was called a 'disorder'. This kind of person is now a moving target. We have (b) the people, those I call 'unhappy', 'unable to cope', or whatever relatively non-judgmental term you might prefer. There are (c) institutions, which include clinics, annual meetings of the International Society for the Study of Multiple Personality and Dissociation, afternoon talkshows on television (Oprah Winfrey and Geraldo Rivera made a big thing of multiples, once upon a time), and weekend training programmes for therapists, some of which I attended. There is (d) the knowledge: not justified true belief, once the mantra of analytic philosophers, but knowledge in Popper's sense of conjectural knowledge, and, more specifically, the presumptions that are taught, disseminated and refined within the context of the institutions. Especially the basic facts (not 'so-called facts', or 'facts' in scare-quotes): for example, that multiple personality is caused by early sexual abuse, that 5 per cent of the population suffer from it, and the like. There is expert knowledge, the knowledge of the professionals, and there is popular knowledge, shared by a significant part of the interested population. There was a time, partly thanks to those talkshows and other media, when 'everyone' believed that multiple personality was caused by early sexual abuse. Finally, there are (e) the experts or professionals who generate (d) the knowledge, judge its validity, and use it in their practice. They work within (c) institutions that guarantee their legitimacy, authenticity and status as experts. They study, try to help, or advise on the control of (b) the people who are (a) classified as of a given kind.

This banal framework can be used for many examples, but roles and weights will be different in every case. There is no reason to suppose that we shall ever tell two identical stories of two different instances of making up people. There is also an obvious complication: there are different schools of thought. In this first instance, there was the multiple movement, a loose alliance of patients, therapists and psychiatric theorists, on the one hand, who believed in this diagnosis and in a certain kind of person, the multiple. There was the larger psychiatric establishment that rejected the diagnosis altogether: a doctor in Ontario, for example, who, when a patient arrives announcing she has multiple personality, demands to be shown her Ontario Health Insurance card (which has a photograph and a name on it) and says: ' This is the person I am treating, nobody else.' Thus there are rival frameworks, and reactions and counter-actions between them further contribute to the working out of this kind of person, the multiple personality. If my sceptical colleague convinces his potential patient, she will very probably become a very different kind of person from the one she would have been had she been treated for multiple personality by a believer.

I would argue that the multiple personality of the 1980s was a kind of person previously unknown in the history of the human race. This is a simple idea familiar to novelists, but careful philosophical language is not prepared for it. Pedantry is in order. Distinguish two sentences:

A. There were no multiple personalities in 1955; there were many in 1985.

B. In 1955 this was not a way to be a person, people did not experience themselves in this way, they did not interact with their friends, their families, their employers, their counsellors, in this way; but in 1985 this was a way to be a person, to experience oneself, to live in society.

As I see it, both A and B are true. An enthusiast for what is now called Dissociative Identity Disorder will say, however, that A is false, because people with several 'alter personalities' undoubtedly existed in 1955, but were not diagnosed. A sceptic will also say that A is false, but for exactly the opposite reason: namely, that multiple personality has always been a specious diagnosis, and there were no real multiples in 1985 either. Statement A leads to heated but pointless debates about the reality of multiple personality, but in my opinion both sceptics and enthusiasts can peacefully agree to B. When I speak of making up people, it is B that I have in mind, and it is through B that the looping effect occurs.

Multiple personality was renamed Dissociative Identity Disorder. But that was more than an act of diagnostic house-cleaning. Symptoms evolve, patients are no longer expected to come with a roster of altogether distinct personalities, and they don't. This disorder is an example of what in my book Mad Travellers (1998) I called a 'transient mental illness'. 'Transient' not in the sense of affecting a single person for a while and then going away, but in the sense of existing only at a certain time and place. Transient mental illnesses can best be looked at in terms of the ecological niches in which they can appear and thrive. They are easy cases for making up people, precisely because their very transience leads cynics to suspect they are not really real, and so could plausibly be said to be made up.

We now read of an autism epidemic and an obesity epidemic, just as we used to read about the multiple personality epidemic (and an epidemic of child abuse). The conception of autism has evolved, as I discussed in the 11 May issue of the LRB . In 1992, The American Heritage Dictionary of the English Language , a reliable desktop dictionary that tries to keep in touch, defined autism as: '1. Abnormal introversion and egocentricity; acceptance of fantasy rather than reality. 2. Psychology : Infantile autism.' In 2000 it gave: 'A psychiatric disorder of childhood characterised by marked deficits in communication and social interaction, preoccupation with fantasy, language impairment and abnormal behaviour, usually associated with intellectual impairment.' The first sense of 1992 derives from the usage coined by the great Swiss psychiatrist Eugen Bleuler in 1908, as the name for the behaviour of some of his schizophrenic patients. The second 1992 sense, 'infantile autism', was a transfer from the first sense and was introduced by Leo Kanner in 1943.

The 2000 definition is about as good as you can do with so few words. It could have added the obsession with literalness, the obsession with order and keeping things the same, the terrible tantrums, biting and hitting that follow when things cease to be the same. It could have added that most people with autism, about four out of five, are male. It could have added the habit of echoing what has been said, rather than speaking. In short, it could have added lots more, but the definition is not bad.

The definition is certainly wrong about one thing: autism is not just a childhood disorder, but almost always for life. It is a developmental disorder that can be recognised very early, usually no later than 30 months, for which there is no known cause and no known cure. At most, it is widely believed, a child can learn to compensate for the deficits, although there are some remarkable recoveries. Another aspect of the definition at which many would protest is its regarding autism as a 'disorder', now the standard euphemism for mental illness. Many advocates for autism insist that it is not a disorder but a disability.

In 1943, indeed in 1973, autism was a rare developmental disorder with a quite definite, narrowly characterised stereotype. Today, we have the autistic spectrum. We have high-functioning people with autism. We have Asperger's, a name introduced into English in 1981 by the British psychiatric social worker Lorna Wing. It is adapted from a diagnosis made in 1944 in Vienna by Hans Asperger, a distinguished paediatrician in the German-speaking world, whom Wing made prominent in English. It now tends to refer to people who had few difficulties acquiring language, but have all the other autistic symptoms. It is often loosely synonymous with high-functioning autism.

Consider a certain kind of teenager or adult, the high-functioning autist. (I shall leave Asperger out of it.) The typical case is someone who grew from an autistic child into an adult who had full or almost full possession of language, and some residual eccentricities of an autistic sort, some of which are socially disadvantageous, some possibly advantageous. Temple Grandin is the most famous example. She emphasises her empathy with animals, urging that her way of seeing the world is closer to that of animals than to most humans. She has had a significant effect on American slaughterhouse techniques. High-functioning autists are beginning to crop up in fiction, much as multiple personalities did twenty years ago. Some high-functioning autistic people talk of forming an autism liberation front. Stop trying to make us like you. We do some things better than you, and you do some things better than us, so leave us be.

Now let's try out A and B for high-functioning autism:

A. There were no high-functioning autists in 1950; there were many in 2000.

B. In 1950 this was not a way to be a person, people did not experience themselves in this way, they did not interact with their friends, their families, their employers, their counsellors, in this way; but in 2000 this was a way to be a person, to experience oneself, to live in society.

As I said, A in my view is true for multiple personality. But it is absolutely false for high-functioning autism. It is almost as absurd as saying that autism did not exist before 1943, when Kanner introduced the name. But B, I believe, is true. Before 1950, maybe even before 1975, high-functioning autism was not a way to be a person. There probably were a few individuals who were regarded as retarded and worse, who recovered, retaining the kinds of foible that high-functioning autistic people have today. But people didn't experience themselves in this way, they didn't interact with their friends, their families, their employers, their counsellors, in the way they do now.

As I see it, this kind of person, in the sense of B, could not have existed until some time after autism itself had been diagnosed, for the first such individuals had to be diagnosed as autistic and then somewhat mysteriously 'recover', grow out of it, acquire social skills, be able to understand what other people are thinking and feeling, overcome, or at any rate live unproblematically with, the obsessive need for literalness. Once there were 'recovered' autists, other adults, who had never been diagnosed as autistic, could be seen as having similar difficulties, even if their childhood was not as bad. Hence the class of high-functioning autists rapidly expanded. Some will have strengths in one direction, some strengths in another.

How does making up people take place? Long ago, 'hip' and 'square' became common names in white middle-class culture. By a parody of Nietzsche, two new kinds of people came into being, the hip and the square. As is the way of slang imported from another social class, both kinds had short shelf lives. But I am concerned with the human sciences, from sociology to medicine, and they are driven by several engines of discovery, which are thought of as having to do with finding out the facts, but they are also engines for making up people. The first seven engines in the following list are designed for discovery, ordered roughly according to the times at which they became effective. The eighth is an engine of practice, the ninth of administration, and the tenth is resistance to the knowers.

1. Count!
2. Quantify!
3. Create Norms!
4. Correlate!
5. Medicalise!
6. Biologise!
7. Geneticise!
8. Normalise!
9. Bureaucratise!
10. Reclaim our identity!

The success of the seven engines of discovery has been astonishing. It is no criticism to say that they have side effects, that they sometimes bring new kinds of people into being, in the modest sense of proposition B, and that they affect the kinds of people they study.

Here are some brief illustrations of what I mean by each of my ten engines. I shall use autism and obesity as contrasting illustrations. They remind us that the ten engines work in different ways on different kinds of people.

1. Counting . The first large-scale, well-designed attempt to count autistic children was made in Camberwell, getting a rate of 4.5 per 10,000, which may still be about right for core autism, as opposed to the extended autistic spectrum. There are now about eighty published countings, and growing, as is the proportion of people with autism. On 4 May, Reuters reported that 'the first national surveys of autism show the condition is very common among US children.' Well, not so common: 57 per 10,000. You will know the horror figures for obesity rates. There is, however, a big difference between autism and obesity, which is nowadays even more assiduously counted than autism. Whether obesity is as bad as it ought to be or not, its prevalence has immensely increased, all over the world, in the past two decades. In contrast, we debate whether the swollen figures for autism show that the prevalence of autism is increasing, or only that we have expanded definitions and are more alert for possible diagnoses.

2. Quantity . In the case of obesity, quantity is built in. We have our bathroom scales. In 1903, the Society of Actuaries and the Association of Life Insurance Medical Directors of America defined being 'overweight' as weighing more than the average for insured people of one's own age, height and sex, and obesity as 'an excessive accumulation of body fat'. During the 1970s, the Body Mass Index took hold, the ratio of the weight of a person in kilogrammes divided by the square of their height in metres. In 1998, the World Health Organisation, in company with numerous national bodies, defined being overweight as having a BMI of more than 25, and obesity as a BMI of more than 30. For a sense of what these numbers mean, Leopold Bloom had a BMI of 23.8. Marilyn Monroe varied between 21 and 24. 'Underweight' is defined as below 18.5. During the past twenty years models in Playboy have gone down from 19 to 16.5. Fauja Singh, the fastest man on earth over the age of 90, has a BMI of 15.4. Autism resists quantity. There are many diagnostic questionnaires, but it's hard to quantify deficits.

3. Norms . Georges Canguilhem's The Normal and the Pathological (1943) showed how medicine acquired the concept of normalcy not long after 1800. Many of our examples are deviations from the norm, for better (genius) or worse (obesity). Canguilhem addressed the question: which comes first, normalcy or deviance? There is no general answer. Sometimes one, sometimes the other, often hand in hand. Quantitative norms followed Adolphe Quetelet's homme moyen in the mid-19th century. To say that autism is a developmental disorder is to say that autistic children do not develop normally. Norms for development ' the ages at which children usually do such and such; can be expected to do such and such; ought to be able to do such and such (tie their shoelaces) ' are a standard part of manuals on child-rearing for parents.

4. Correlation . This is the fundamental engine of the social sciences. It began around 1870, when Francis Galton devised the correlation coefficient. Quetelet had the mean, but Galton made deviation from the mean the core of his social philosophy, and so devised the correlation coefficient. We try to correlate autism with everything, not excluding the relative lengths of the mother's fingers and amount of testosterone in the foetus. Some correlations need no statistical theory or analysis: four out of five children with autism are male. On the other hand, bodyweight needs subtle statistics. A BMI between 25 and 30 is said to be bad for you because of a significant correlation with numerous risk factors, which are themselves statistical entities. It is a strange situation. Being overweight, unlike being obese, does not importantly affect your life expectancy, but unless you are a body builder or rugby forward, it will make you less attractive in current society, less physically active and so forth.

5. Clinical medicine . We medicalise kinds of deviant people relentlessly, not always with success. The modern concept of child abuse was introduced by doctors around 1960, but there have been substantial battles over the 'medical model' ever since. There have always been fat people, some of them ill. But stout, plump persons have often been in fashion, as the works of Rubens or Renoir remind us. 'Let me have men about me that are fat, sleek-headed men and such as sleep o' nights.' Today, we treat the stout as having medical problems, and the obese as needing medical instruction. A new generation of anti-craving medicines is about to make a fortune for its owners. Autism was regarded as a diagnosis made by a child psychiatrist, and so it is filed as a mental disorder and hence in the end as a medical problem. But if we regard it more and more as a disability, it may seem less and less medical.

6. Biology , including neurology. Autism is a disability but it has biological causes, specifically neurobiological. One of the great moral benefits of biologising is that it relieves a person of responsibility. If overeating is attributed to chemical imbalance it ceases to be a moral defect.

7. Genetics . There is now a constant drive to trace the medical to the biological, and the biological to the genetic. This is not wholly new. A century ago, there was a great push to discover the genetic origins of criminal behaviour, of the criminal personality and so forth.

8. Normalisation . In many cases, we try to make unfavourable deviants as close to normal as possible. That is the point of the behavioural therapies for autism, and of anti-craving drugs for obesity. A perspective different from mine would emphasise that this is where all the action is. It would urge that ideas do not change people. Treatments do, whether behavioural or pharmaceutical.

9. Bureaucracy . Some schools of thought speak of bureaucratic power as if it were always a bad thing. So let me emphasise the positive. Most prosperous nations have quite complex bureaucracies that pick out children with developmental problems in the early years of schooling, and assign them to special services. The system sees itself as an objective way to determine who needs help, but the relation is reciprocal. The criteria used by the system in turn define what it is to fall under various categories such as autistic. There is an ongoing feedback effect. Once again, obesity is a contrast case, for it has not yet in any important way been bureaucratised.

10. Resistance . Kinds of people who are medicalised, normalised, administered, increasingly try to take back control from the experts and institutions, sometimes by creating new experts, new institutions. The famous case is homosexuality, so highly medicalised from the time of Krafft-Ebing late in the 19th century. That was the very period in which legal institutions became active in punishing it. Gay pride and its predecessors restored to homosexuals a control of the classifications into which they fall. There are always twists and turns in the tales of making up people, few more striking than the attempts to geneticise male homosexuality, to find the gay gene.

I mentioned motions towards an 'autism liberation front', something that would make high-functioning autistic people the experts on their condition. There are a number of organisations of overweight and obese people trying to reinstall pride and dignity in heavy bodies. I like, both for its acronym and its activities, a rather modest and cautious French organisation: Groupe de Réflexion sur l'Obésité et le Surpoids, or GROS.

All ten engines, including the seven engines of discovery, produce effects on the kinds of people to whom they are applied. They change the boundaries. They change the characteristics. My names and my sorting of modes of inquiry may be eccentric, but they are readily recognisable. Conjectures about obesity and autism abound but, fortunately, there is competition. Different groups make different guesses about which will be corroborated. We might find that there is no genetic basis for autism, and none for all but a small proportion of obese persons. Or we might find that most obesity and all autism are linked to certain organisations of genetic anomalies. It is important to know. We find out thanks to the seven scientific engines, and we usually end up knowing more than we did before. I observe only that we tend to think of the engines as directed at fixed targets. The targets, however, do not stay still. This in no way calls into question the objectivity of the engines of discovery.

There may be, in the genetic make-up of human beings, a rather rare set of genetic anomalies that is responsible for most cases of autism. If so, it is a fixed target at which we aim, although we don't know what it is. The anomalies (if there are such) cause a person to be autistic. They do not determine the ways of life for autistic people. If we came to know such a set of anomalies, they might even redefine autism, excluding children who are thought to be autistic because of their behaviour. Think about the ways that the disability we call autism has changed its contours and its lived experience during the past sixty years. That is the moving target.

Often when we try to generalise we go into the species mode. Indeed, there are books called The Autistic Child and The Obese Child . But some autism advocates strongly object to speaking of 'the autistic child' and prefer 'children with autism'. One can sense what they are opposing. To speak in the species mode about people is to depersonalise them, to turn them into objects for scientific inquiry. Other thoughtful people feel that 'autistic child' is just right. A parent who founded the Autism Society of America, and wrote one of the first books about the topic, did so because 'autism is who my son is, not just a characteristic.' It is part of the boy's nature to be autistic. Except in very rare cases, I am disinclined to say the same thing of an obese person. Being overweight is always just a characteristic: it is never who the stout man is, just one of his enduring, and maybe endearing, properties.

John Stuart Mill, the progenitor of the doctrine of natural kinds, left us a good way to distinguish the two. Giving horse and phosphorus as examples, he argued that there are endless characteristics associated with some classifications; thus horses (and phosphorus) have innumerable features in common, in addition to their being horses (or phosphorus), while white things have nothing much in common except that they are white. Horse , he wrote, was a 'real kind' (of animal), what philosophers later came to call a natural kind. White was merely a finite kind. He worried about whether the races and sexes were real or finite kinds, and opted for members of the different races having no more in common than their race, just as Christians have nothing in common except their faith. The races and sexes are therefore not real kinds.

Mill's distinction expresses well the idea that 'autism is who my son is, not just a characteristic.' Autistic children have a wide range of characteristics in common, distributed on a spectrum, or, I prefer to say, in a space that is at least three-dimensional. Some of these are built into diagnostic interview schedules for detecting aspects of the autistic spectrum. Others, less behavioural, are unknown, and are, so far, hidden in bio-neuro-genetic space. This is not true of overweight people. They have little in common except that they are rather plump. It is not true of obese people: they have nothing in common except that they are fat. There may be subclasses of obese people who have a distinct biological cause for their having a BMI in the very high range. Whatever it is may be part of their nature, and may bring in a host of other characteristics. That subclass would come close to being what Mill called a 'real kind'.

A vigorous school of cognitive science argues that the tendency to treat kinds of things and kinds of people as if they had essences is innate in the human mind. It does not claim that there are metaphysical essences: only that we innately think and act as if there were. I am sceptical. Perhaps this tendency might be better studied under the heading of the historical anthropology of scientific reason. The cognitivists will protest that their results are confirmed cross-culturally and apply to six-year-olds. Well, yes: to six-year-olds who grow up anywhere in the world of scientific reason, what Marshall Sahlins calls 'the world system'.

It is part of our scientific attitude that what we find out about people using any of the seven engines of discovery, and more, is a fixed target. We usually hit something, and then we say that what we hit was what we were aiming at. What we find out about it is for the most part true, or not far from the truth. Yet the target that we hit is often where it is because of the interaction between our five elements, ranging from classifications through people to experiments. Sometimes this breeds conceptual confusion. There may be no better example than the changing faces of suicide.

Suicide is now tied to depression. 'An attempted suicide is a cry for help.' Nothing is more shattering than the suicide of a friend. Nothing more smashes the spirit of a psychiatrist than the suicide of a patient. Nothing seems more awful than for young people to kill themselves. When a wave of suicides passes through an adolescent cohort in a native village in northern Canada, well-meaning citizens in the south are steeped in shame and guilt. This wholly modern feel to suicide, and the gamut of associated meanings, is a product of interaction with statistical and medical sciences, a family of interactions that began around 1825. This modern arrangement of intense feelings and meanings makes us totally confused when we think about either euthanasia or suicide bombing.

The latter is a ruthless weapon, often callously exploited by older men who have no intention of killing themselves. It is nevertheless a remarkable response by angry young Muslims who feel impotent when faced by omnipotent hegemony. It can be used by anyone: Tamil Tigers developed much of the early technology. The suicide weapon is the polar opposite of the invincible nuclear weapon. But they are a match, equally indifferent to whom they kill. We have great difficulty thinking about the suicide weapon because of our established scientific knowledge about suicide. That knowledge is nevertheless true knowledge about people in our society, the suicides and those who meditate self-destruction. They have grown through their lives to conform to the meanings and the stereotypes that the knowledge teaches.

Genius has put on an amazing number of masks since the word was used with such effect in antiquity. The term ' I hardly dare to call it a 'concept', but perhaps one could say 'cluster of associated ideas' ' maps the fantasies of the age: Athens in its prime, Elizabethan England, Romantic Germany, Fin de Siècle France. But genius is not a serious concept in our times. It has quite lost the allure of the Romantic era. That is because we now measure it, and genius of its nature abhors a measure.

Starting with Galton's Hereditary Genius , we have gradually made intelligence statistical, with norms. Indeed, the usual IQ tests are so statistical that the questions are designed in such a way that a curve of scores forms a normal distribution with a mean of 100. When the tests were first applied to women, they scored higher than men, with a mean of about 105, so the questions had to be modified to make them harder for women. They were adjusted until the mean score for females was also 100.

IQ tests are excellent at evaluating the ability of a numerate and technical child, possessing a new kind of literacy, to prosper in our times. At the top end, genius is forced onto a linear scale and hence off the map. Galton aimed to measure genius but in fact he expelled it from our culture. It is part of the deep, ultimately Socratic notion of genius, that when it is measured on scales that stem from Galton, and were refined in 1917 by the United States army for evaluating recruits, true genius ' I don't hesitate to use that phrase ' will be living somewhere else. It will blithely refuse to interact with questionnaires, institutions, experts and knowledge, rejecting classification. Ah, as you see, I have just bought into the Romantic notion of genius.

Ian Hacking holds a chair in the philosophy and history of scientific concepts at the Collège de France in Paris.

Published in the LRB Vol. 28 No. 16 dated 17 August 2006

little site banner