What is public health? Since 1928, when the question was first asked to the participants of a symposium held by the American Public Health Association, it has received many answers. The purpose of this paper is not to propose an additional definition, but to analyse how public health is socially constructed and produced. Instead of prescribing what it should be, I will describe it as it is. More precisely, I want to show that it is a human activity which transforms natural phenomena into cultural facts. In order to defend and illustrate this point, I will present a case study on the history of lead poisoning in France.
In 1981, a team of paediatricians at the Edouard Herriot University hospital in Lyon, with toxicologists from the Alexis Carrel Faculty of Medicine, published an article entitled ‘Lead poisoning revealed by severe encephalopathy: pica does exist in France’ in the journal Archives françaises de pédiatrie1. The article examined the case of a 5-year-old boy called Mammar, hospitalized for consciousness disorders and digestive symptoms. While the diagnostic investigations dragged on without conclusive results, the child’s neurological condition worsened. After several convulsions, a neurosurgical operation was carried out on the assumption that he may be suffering from hydrocephalus. A few days after the surgery a blood test eventually revealed lead poisoning, confirmed shortly afterwards by an interview with his parents who said that their child ‘ate the paint flaking off the walls of their dilapidated house’. Lead poisoning had nevertheless been difficult to identify because in France at the time it was thought to be ‘rare in general pathology, especially in paediatrics’. The child was treated by chelation and his clinical condition improved. Six months later, we are told, he had fully recovered and had no after-effects. Yet no social investigation was carried out: there was no visit to his home, his siblings were not given blood tests, and the family was not offered alternative housing. The little boy simply went home and most probably carried on poisoning himself, if not by eating flakes of paint, assuming that his parents kept him away from that source of contamination, then at least by inhaling lead particles, which would have been unavoidable. At the time, doctors saw lead poisoning as a clinical problem the treatment of which was entirely the responsibility of the hospital. Mammar’s sickness was thus destined to remain a ‘success story’, especially since he was not one of the dramatic cases of encephalopathy ‘resulting in death (in 30% of cases) or in mental backwardness or convulsions (in 80% of cases)’, as the authors commented. Between 1956 and 1981, only 10 cases of infantile lead poisoning had been reported in France.
Less than two decades later, in 1999, an inquiry by Inserm, the French National Institute for Health and Medical Research, estimates the number of child victims of lead poisoning at around 85,000, corresponding to a contamination rate of 1.94% if the international threshold of 10 μg/dl is applied2. The report proposes two types of protocols for large-scale screening and intervention. The first one, a population approach, implies a screening of children from low-income families and, once they are diagnosed as contaminated, investigates the housing in which they live. For the 50,000 homes thus identified, the ‘cost of palliative work is estimated at 2.5 billion francs, and 5 billion francs if the lead is to be eradicated’. The second one, an environmental approach, focuses on old and dilapidated housing supposed to expose its occupants to high risk of lead intoxication. For the 150,000 homes selected through this procedure, ‘the cost of the strategy is estimated between 11.5 and close to 20 billion francs’, the authors assert.
Between the 1981 article and the 1999 report, the isolated case has thus become an epidemic. In addition to clinical identification, the goal is now also mass diagnosis. The treatment of children by chelation is being backed-up by treatment of their homes by renovation. A disease that was said to be exceptional has finally become a political priority inscribed in a law passed in 1998. Lead poisoning has entered the public health domain. Meanwhile, however, there has been no spectacular increase in the prevalence of infantile lead poisoning. It is the way in which the disease is viewed that has changed3. It is now an issue of population, risk, thresholds, collective measures instead of individuals, symptoms, biology and medicines. The fact that the child is called Mammar, i.e. is of African origin, that he comes from an underprivileged home and lives in a dilapidated building, is now important, whereas in the past, these were simply anecdotal elements in a clinical case.
This new set of words and things, signs and facts, that replaces or shifts the medical set can be viewed as the expression of a public health culture4. It is a social construct that generates meaning through which the world is represented, here in terms of an epidemic that calls for strong political mobilization, rather than in terms of isolated cases left in professionals’ hands. In other words, it can be defined as ‘an historically transmitted pattern of meanings embodied in symbols’5. But childhood lead poisoning is not only the result of a translation from clinical language into epidemiological facts, it is also a disease produced by insalubrious housing; public health significations are not only incorporated into cultural symbols, they are also embodied in physical organs. Beyond social operations and closely linked to them, lead poisoning imposes its materiality inscribed in bodies6. The first perspective which focuses on the role played by social agents to establish the evidence of the epidemic is classically considered as constructionist. The second perspective which emphasizes the way in which the social structures produce inequality in the distribution of lead poisoning can be called realistic. In order to analyse this culture of public health and to understand how it relates to the nature of the problem, two hypotheses will be proposed on the naturalization of its objects and the culturalization of its subjects.
The first proposition is that public health naturalizes its objects. This operation, which consists of inscribing evidence in the nature of things, generally eludes analysis, since it seems so obvious—and, precisely, so natural. Yet it is an essential feature of any medical culture, although possibly the most hidden one or, more precisely, the least questioned7. Today, for the journalist or government minister, for the organization volunteer or municipal doctor, the 85,000 child victims of lead poisoning exist as such an obvious fact that they all forget to question the previous evidence of a rare disease. Health problems are not only biological realities that specialists elucidate, they are also epidemiological facts that they construct8. Public health is not content simply to discover—it invents.
To revert to our original children poisoned by lead. In 1981 in Lyon, the little boy had been sent back to his parents and the doctors had considered it enough to see him for consultations, as was standard practice in French paediatric units. In 1985 in Paris, a little girl who was less severely poisoned however aroused more concern among the doctors treating her. They went beyond the usual prerogatives of their profession and organized a visit to her home9. There the welfare worker noted that the child’s family, of Malian origin, was living in a dilapidated flat with flaky paint and floors that were caving in. Dismayed by what she saw, she alerted the municipal maternal and child health services, but had difficulty convincing them of the gravity of the situation. Nobody at the time thought that lead poisoning could be a public health problem10. The social paediatrician who was in charge of the area later admitted that she had only ever heard of the disease in connection with occupational medicine and was totally unaware that it existed in children.
But, from that moment, lead poisoning had left the confines of the hospital. Children were still diagnosed and treated in paediatric wards, but as more and more clinicians tended to recognize this reputedly silent and polymorphous disease, other agents—e.g. public health doctors, toxicology specialists and social workers—increasingly took an interest in it as well. Recognition of the problem gathered momentum. A retrospective study identified 20 cases in Paris between 1984 and 1986, and two small-scale epidemiological studies were undertaken in 1986 and 1987 by a doctor and a biologist, both with little experience in biostatistical research11. One study was descriptive: it covered the 52 residents of two buildings in which the first cases had been diagnosed. It demonstrated that only children under the age of six showed signs of poisoning, in a proportion of four to seven. The findings excluded contamination through water, which had initially been incriminated, and even the researchers were surprised to discover that the paint on the walls contained extremely high levels of lead. The second study took an aetiological approach. It compared 82 children at risk because they lived in dilapidated homes, with 40 children living in newer buildings, both located in a similar social environment. The causal role of the housing conditions was established, as seven cases of poisoning were found in the first group and none in the second, whereas in the apartments of the contaminated children high levels of lead were found on the walls and woodwork. From then on the movement spread: the maternal and child health services set up targeted screening; the public health department of the Bichat medical faculty proposed a monitoring system; and two humanitarian organizations arranged a mission to the USA to investigate the screening and prevention measures implemented in American cities. In 1990, 5 years after the first case that had stirred people into action, 1500 cases of lead poisoning in children had been recorded in Paris alone12. An epidemic was born.
For this translation of clinical observations into epidemiological data to be possible, a change had to occur in the worldview of medical professionals—or at least of some of them13. First, new intellectual objects had to be mobilized. There was a gradual shift away from the language of symptoms, doses, X-rays and chelation, towards groups, screening, risks and health measures. Individual reasoning based on clinical observation made way for a population-oriented approach to public health. Second, new practical tools were applied: in addition to treating contaminated children, surveys were undertaken on the residents of dangerous buildings; professionals started to calculate poisoning rates rather than simply relating the cases of contaminated children; and evidence was gathered through comparative studies rather than by simply questioning parents regarding their children’s habits. With these objects and tools it became possible to establish the frequency of the health problem, its target and its cause, all of which had formerly been impossible to infer simply by reading the North American medical literature on the subject, despite its availability.
This change of perspective was of the same nature, although on the scale of the small Parisian medical world and a much shorter time-span, as the one that had spawned ‘the rise of statistical thinking’ a century earlier14. The main difference was that knowledge through figures was now socially and scientifically accepted. It was no longer a question of imposing it against the supporters of the individual approach, as the promoters of the ‘numeric method’ had done15. But it had become necessary to apply an interpretative framework that French doctors were ill-equipped to use. Their training included public health based on programmes inherited from the great treatises of the late 19th century, but no initiation into statistics16. In the early 1980s in French medical faculties, epidemiology was still a description of epidemics, not a science of numbers. It was only under the two-fold influence of an academic current of mathematics applied to health, and an activist movement demanding a community approach to medicine, that the ‘new epidemiology’ made its way into universities, districts and ministries. But until this point it was still a matter of scholarly culture.
In fact, the social construction of the lead poisoning epidemic must be set in a vaster context of cultural change through which the governance of human beings affects not only individuals but populations, not only care for the ill and, more generally, responses to demands, but also risk prevention and, more broadly, anticipation of problems17. In the case of lead poisoning in children, resistance to this new approach was strong. The Paris municipal housing authorities dragged their feet when it came to rehousing, and balked at participating in inquiries on old buildings. National health authorities were reluctant to recognize prevention of lead poisoning in children as a health priority. Until the 1990s those agents working towards and calling for recognition of the reality and gravity of the situation were marginal and lacked legitimacy, as a result of either their disciplinary affiliation (in the case of local public health doctors), their institutional position (in the case of members of humanitarian organizations) or, finally, their profession (in the case of social workers). Even these agents themselves sometimes seemed unsure of the validity of their struggle. Yet this denial is not peculiar to France’s history of the disease, as the North American precedent shows18. Although it sometimes reveals conflicts of interest of a political or economic nature, it attests, above all, to the slowness of the processes through which a common public health culture is formed.
But in order to understand fully the process of construction of childhood lead poisoning, one must take into account a reality rarely analysed as such: it is the fact that, by transforming cases into an epidemic, the social agents tend to describe a different disease. Indeed this is true every time distinct medical worlds describe their common object in a different manner19. The lead poisoning reported by paediatricians only 10 times in a quarter of a century is not the same disease as the one for which public health specialists counted 1500 cases in 5 years. The former was generally manifested in severe brain damage, a large proportion of which we can reasonably assume was never recognized as such. The latter usually corresponded to asymptomatic forms, discovered through interviews oriented towards housing conditions. To be sure, in both instances it was a matter of an accumulation of metal in the body, but even this physiopathological fact has changed. In the previous situation, levels of lead in the blood often exceeded 100 μg/dl, whereas in the new configuration, six out of seven cases of contamination were under the severity threshold which, at that time, was set at 50 μg/dl. The social agents noted that since screening had started, no more stories of acute encephalopathy had been recorded. In other words, as the epidemic grew, the number of clinical cases paradoxically declined.
Redefinition of the health problem is most clearly evident in the repeated evaluations of the toxicity threshold. When paediatricians related their alarming observations in the early 1980s, they set normal lead levels at 35 μg/dl, according to medical literature of the time. As the two local epidemiological studies were carried out later in the decade, the norm had been empirically shifted to under 25 μg/dl. By 1990, the 1500 Parisian cases were identified on the basis of the Centres for Disease Control limit of 15 μg/dl. Later on, following North American surveys that established a correlation between lead levels considered to be low, and cognitive and behavioural disorders in children and adolescents20, the internationally recognized value became 10 μg/dl: it is in relation to this value that the number of contaminated children is estimated at 85,000 by the Inserm report. Logically, as the threshold declines, the epidemic swells in statistical terms. The decision to adjust the norm was obviously not an arbitrary one since it was based on improved knowledge of the long-term effects of lead. Yet it ended up producing a new disease: encephalopathy as a result of lead poisoning that a good clinician ought to be able to identify, has been replaced by a reality that only calculations of probability can grasp (provided that social variables are checked), and that is manifested in a higher risk of backwardness at school and acts of delinquency (in statistical terms). It is another way of making sense of the pathology thus defined by public health. While claiming to reproduce a natural reality, it transforms it into a new object, unrecognizable but politically relevant. The idea is no longer to treat some children better, but to renovate entire neighbourhoods. However, the public health culture is manifested not only in its way of defining the disease but also in its way of considering patients.
The second proposition is that public health culturalizes its subjects. In other words, it produces statements and acts on the culture of those for whom it is intended and whose representations and practices it is designed to change so that they may have a better or longer life. It thus institutes a relationship of otherness with its public: it has the knowledge, the other has beliefs21. Taking care of public health means changing the latter to bring them closer to the former—an operation even more effective when the other is socially, ethnically and geographically distant, in other words, when it appears almost naturally in its cultural difference22. The poor, the immigrant, the youth, the Indian woman or African peasant are all subjects who, despite themselves, tend to be culturalized. This applies to any public action, but because it comprises a strong moral dimension of acting for the good of humanity, public health is probably exposed more than other areas.
Consider France’s history of lead poisoning. In 1981 in Lyon, nothing had been said about the little boy’s origins. Four years later in Paris, we are told that the little girl is African and that she lives in a dilapidated flat, like the five other cases diagnosed in the same hospital during the next few months, and like the children diagnosed before 1990: all were foreigners, 14% from North Africa and 85% from sub-Saharan Africa. “In Paris and its surroundings, most lead poisoned children are of African origin and black”, noted the team of toxicologists23. Yet their explanation remained uncertain: “The ethnic factor is correlated with unfavourable socio-economic situations, including bad living conditions. Moreover, pica behaviours are more easily tolerated in families where geophagy is a cultural element”24. Pica is a distortion of the taste for food that causes people to ingest mineral substances and, for a long time, it was the most common explanation for lead poisoning in children contaminated by eating paint flakes off walls and wood. It was combined here with geophagy, a practice of eating dirt assumed to be characteristic of West African peoples and believed to induce a tolerance for children’s pathological behaviour. Lead poisoning was thus considered to be the consequence of a cultural habit inducing pathological behaviours. In reality, from the outset the children’s African origins had intrigued public health specialists.
How could one account for this massive over-representation of Africans from West Africa? When questioned on their profession, many fathers had mentioned that they were ‘marabouts’, i.e. Muslim healers—an activity that often serves to mask the reality of joblessness. Some professionals soon suggested they were dealing with “a disease of marabouts’ children” and looked for a possible source of contamination in the ink of Muslim prayer tablets. More broadly, African practices and objects were closely examined in interviews with parents and compared with lists of dangerous products, before being subjected to chemical analyses: eye-shadows, craft pottery for cooking food, and traditional potions administered by parents25. Faced with the authorities’ incredulity as to the reality of the epidemic and its source, it was necessary, they explained, to eliminate all possible causes. Nevertheless, this excess of cultural zeal is astonishing if we consider that in North America, for more than half a century, paint had been considered as the main source of poisoning of children, and that tests on paint during the first Parisian inquiry had shown very high lead concentrations. However, it was not long before the role of old paint became difficult to deny. This was when pica came onto the scene.
If dilapidated housing conditions are a concern, why are all the victims African? If the reasons are not genetic, they must be cultural, we were told. The first interpretations were common sense: the children lacked toys and stimulations, they were bored and spent their time on the floor or in front of a window, left to themselves; the mothers were mostly uneducated and ill-informed, they failed to watch their children or to check their deviant appetites. This was nothing new, for in the 1960s in the USA, this reproving discourse against Black women in poor neighbourhoods, accused of not caring for their children properly—with its cultural explanation to tone down its sententiousness—was already widespread26. Soon these analyses gained academic status in France with the arrival of a couple of anthropologists who embarked on a quest for ‘specific socio-cultural characteristics over and above the main risk factors, related to the over-population of precarious, dilapidated and run-down housing’. These specialists of the ethnology of West Africa discovered this characteristic in ‘natives’ habits of accepting geophagy’ leading ‘to a particular tolerance of the sight of a child sucking fragments of wall-covering’27. In terms of a model borrowed from ethnopsychiatry, pica was distinguished from geophagy: whereas the latter was seen as culturally normal and codified within this context, the former appeared to be a pathological variant that the Africans themselves recognized as such.
While it did not deny the responsibility of paint and the reality of poverty, this interpretation—as many activists noted with indignation—nevertheless shifted the focus towards a collective explanation, culture, and an individual one, behaviour. It finally justified health education that, for economic reasons, was more and more often proposed instead of re-housing, i.e. asking mothers to cut their children’s nails and to wash their floors often28. Yet history would have had a lot to teach these experts in otherness. In the 1940s, seminal research in Baltimore showed that five times more black children died of lead poisoning than white children, even though there were four times as many white children, and suggested that irrespective of how good their intentions were, educational programmes were largely ineffective in those groups who were well aware of the danger of lead but were financially unable to move away from it29. The answer to the very relevant question of why there is a high incidence of infantile lead poisoning among African children today in Paris and its surroundings, resembling the same high incidence among black children in Maryland 60 years ago, obviously needs to be sought elsewhere, not in culture. The following historical interpretation also serves as a political analysis.
In the early 1970s, French society officially closed its borders to migrant labour. Family reunification was only allowed for a few more years30. At the same time the economic slump, coupled with rampant unemployment, resulted in drastically reduced social and residential mobility. Within a few years, Africans from sub-Saharan Africa and primarily from the former French colonies, who constituted the last large wave of immigration, were cut off from the job market and from access to social housing. Faced with economic and even legal precariousness, they had to settle for the most dilapidated housing in which they were sometimes squatters, or that was rented at high rates that they were not in a position to refuse. In some cases, this situation was worsened by the arrival of several wives and large families. These people currently account for a large proportion of the inhabitants of most underprivileged inner city areas or suburbs of large cities. The fact that under these conditions childhood lead poisoning affects them far more than others is the result of a history illuminated by the political economics of immigration, far more than the consequence of cultural attributes. Yet it is remarkable that in explaining pathologies and contexts as diverse as tuberculosis in South Africa31, Aids in Haiti32 or malnutrition in Brazil33, public health has so often been inclined to investigate the latter rather than questioning the former. This repetition of similar scenarios attests to the permanence of a form of practical culturalism that essentializes culture and makes it a last resort interpretation of these inequalities34. To be sure, not all cultural approaches to public health can be interpreted in this way, but practical culturalism is indeed a dominant feature.
Public health is generally presented as action-oriented knowledge, expertise at the service of decision-making. The materials fed into this knowledge and expertise are considered to be facts of nature that simply have to be made visible in statistics and in the language of determinants and risk factors. The supposed beneficiaries of actions and decisions are assumed to belong to a culture that has to be transformed for their well-being. This separation between two worlds, supposedly objective and subjective, natural and cultural, lies in the foundations of public health, irrespective of variants of and differences within it. Taking it into account therefore implies, as it would for medicine35, that we restore the tension between the material facts of diseases, suffering, pollution, and the social operations of the formulation of rates, the search for causes, and the implementation of programmes.
Lead poisoning affects children, mostly from African immigrant families living in insalubrious buildings. Depending on the threshold chosen, it can be described in terms of serious neurological problems or of a statistically significant decline in the intelligence quotient. Depending on one’s perspective, it is either a disease with cultural origins or the outcome of immigration and housing policies. These choices will impact on the way in which the health problem is addressed. For example, the lower the norm, the higher the estimated number of cases will be, so that a mobilization of resources will seem justified. If the disease is seen as the result of behavioural problems, the health education approach will seem appropriate, whereas if it is seen as the consequence of social inequality, structural measures will be advocated.
Public health is thus the cultural activity through which a biological fact, here lead poisoning in children, is constructed as a social fact, an infantile epidemic with its figures and images, its economic and ethnic characteristics, its aetiological models and its practical answers. ‘The social construction of what?’, one may ask36. Are we challenging the reality of the phenomenon, or its gravity, or its mechanisms? Considering public health as a culture does not mean adopting a relativist standpoint. Social reality is as real as biological reality. One could even say that the former is the weapon that can be used to act upon the latter. We simply need to be aware of that if we are to be adequately prepared for battle.
This text is based on a study which was funded by the French Ministry of Research and conducted in collaboration with Anne-Jeanne Naudé. It has been translated by Liz Libbrecht.
Arch Fr Pediatr
Plomb dans l’environnement. Quels risques pour la santé?Paris: Editions Inserm,
The Birth of the Clinic. An Archaeology of Medical Perception. New York: Vintage,
Critique de la santé publique. Une approche anthropologique. Paris: Balland,
The Interpretation of Cultures. New York: Basic Books,
The Woman in the Body. A Cultural Analysis of Reproduction. Boston: Beacon Press,
J Public Health Policy
Une politique urbaine de santé publique. Les débuts du saturnisme infantile en France. Bobigny: Cresp,
J Toxicol Clin Exp
Revue de pédiatrie
Bull Epidemiol Hebd
The Rise of Statistical Thinking 1820–1900. Princeton: Princeton University Press,
La politique des grands nombres. Histoire de la raison statistique. Paris: La Découverte,
Précis d’hygiène et de médicine préventive. Paris: Baillière,
Risk Society. Towards a New Modernity. London: Sage,
Deceit and Denial. The Deadly Politics of Industrial Pollution. Berkeley, CA: University of California Press,
Differences in Medicine. Unravelling Practices, Techniques and Bodies. Durham and London: Duke University Press,
N Engl J Med
Demystifying Mentalities. Cambridge: Cambridge University Press,
Infections and Inequalities. The Modern Plagues. Berkeley, CA: University of California Press,
Am J Public Health,
2004; in press
Chroniques du saturnisme infantile 1989–1994. Enquête ethnologique auprès de familles parisiennes originaires du Sénégal et du Mali. Paris: L’Harmattan,
Brush with Death. A Social History of Lead Poisoning. Baltimore: The Johns Hopkins University Press,
The social construction of illness is a major research perspective in medical sociology. This article traces the roots of this perspective and presents three overarching constructionist findings. First, some illnesses are particularly embedded with cultural meaning—which is not directly derived from the nature of the condition—that shapes how society responds to those afflicted and influences the experience of that illness. Second, all illnesses are socially constructed at the experiential level, based on how individuals come to understand and live with their illness. Third, medical knowledge about illness and disease is not necessarily given by nature but is constructed and developed by claims-makers and interested parties. We address central policy implications of each of these findings and discuss fruitful directions for policy-relevant research in a social constructionist tradition. Social constructionism provides an important counterpoint to medicine’s largely deterministic approaches to disease and illness, and it can help us broaden policy deliberations and decisions.